Original subtitles

[Music]

thank you

[Music]

welcome to the ccnp voice C voice exam

video certification Series this is

module one and we're going to be going

through a course introduction and a lab

topology overview

let me introduce myself my name is Josh

Kittle or you may know me from Twitter

as Cisco voice dude I've been working

with Cisco unified Communications pretty

much since the beginning you know the

call manager 3.x days and I'm currently

employed by a Cisco partner working in

the delivery side of the house so I

install and build and upgrade and patch

and maintain these types of systems

pretty much every day

um I got my CCNA in 2002 and I'm

currently a ccnp voice and I'm in

process of studying for my ccie

collaboration we're just waiting on

November to come around when the the

tests become available so I've got a

little bit of a background in this stuff

and uh you know I've kind of been there

done that seen lots of different things

but what's cool to me is you know every

new day every new version of software

you know is a new feature is a new

opportunity to learn and that's really

why I you know I kind of do the voice

thing but anyway

um enough about me

um we're going to go through the video

course here

and if you haven't seen videos that I've

done in the past

um you know my style may be a bit new to

you but you know there's going to be

some lecture there's going to be some

presentation but there's going to be a

lot of interactivity going on I'm going

to be asking questions telling stories

and demonstrating things so hopefully

this uh the style is something that you

find useful I'll certainly do my best to

give you the best training experience

possible if you're watching this video

you're doing it probably for one reason

you're going to go after the Cisco ccnp

voice certification and you know this is

one of the videos or I'm sorry one of

the courses that people tend to start

with and you know do first it's a good

foundational concept type of exam

first things first cisco.com Go slash

certification is going to be a resource

that you're going to find useful and you

know even if you don't find it useful

you're going to find yourself going

there anyway and utilizing it as a

reference point throughout your

certification process when looking at

the Cisco ccnp voice and really what

Cisco wants you to take away from the

certification or kind of their synopsis

if you will of the of the certification

is that it's going to validate that you

have the advanced knowledge and the

necessary skills required to integrate

unified communication Solutions into

underlying Network architectures now

those might be new architectures or they

might be existing architectures that

you're going to then you know voice or

collaboration enable

they want to make sure that you're

capable to implement operate configure

and troubleshoot a converged unified

Communications Network and you know

we've gone through different

um you know different explanations or

different buzzwords if you will over the

last decade and a half for you know what

we're now calling unified Communications

environments you know Cisco had the

phrase avid and then you know we've

called things voice over IP but really

unified Communications or unified

collaboration is kind of where we are

today and thanks so we want to make sure

you're able to you know deal with these

Technologies and it focuses the

certification focuses on Cisco unified

Communications manager it's going to

focus on knowledge of Ip quality of

service on the infrastructure it's going

to focus on Gateway configuration

Gatekeepers to some extent IP phones

voice applications and utilities on

Cisco routers and switches and really

it's bigger than that you know we're

going to get into you know the servers

that run the stuff and everything else

but Soup To Nuts Cisco unified

Communications that's that's kind of

what the ccnp voice is all about

you've seen the certification triangle

before

um you know everybody has their own

twist on things so here is mine when

you're going after the ccnp voice you're

likely going to be starting with the c

voice video and we're going to assume

that you've got some background and

let's take a step back for a moment and

talk about what that background

expectation is and really you know the

course that you need to follow or the

sequence that you need to follow in

order to obtain your ccnp voice so it

all starts with the CC ENT and CC ENT is

what we you know what we kind of think

of as the foundation of everything the

CC ENT is Cisco's entry level exam it's

even you know a precursor or you know a

what do you want to call it a

prerequisite for obtaining The Cisco

CCNA certification now keep in mind

to get the CCNA voice certification

which comes next you don't have to have

a CCNA and that's that's something

that's changed you know originally you

did

what's required now or these days is you

require your CC ENT now I'm going to

assume that you're going to go after

your CCNA I'm going to recommend that

you go after your CCNA because it's

going to give you valuable Foundation

knowledge that is going to truly help

you as you you know proceed in your

career and with your studies so you know

we've got the first tier out of the way

now you're into CC Naval Ace and we're

talking the Icom 8.0 exam if you haven't

already watched it that we've put

together a nice video series on that and

I think we did a pretty pretty good job

of preparing you for the exam with that

and hopefully can live up to that with

this series as well once you're into the

ccnp voice Arena you know you've got

five exams that are going to be

necessary to pass to wrap up your

certification C voice which is you know

unified Communications fundamentals

gateways quality of service Etc which is

this series here it's

642-437 as of this recording you'll go

into cipt1 and cipt2 those two exams

cover unified Communications managers

cipt1 is more common is your

fundamentals and Core Concepts in a

single site environment and then cipt2

gets into multi-site stuff gets into

more advanced security stuff

Etc t-voice is a troubleshooting class

it's going to test your ability to you

know really focus on and resolve

problems as they occur in The Voice

network so you're going to learn to read

debugs and things like that and

understand how to break apart various

problems and locate The Faults and cause

them you know and resolve them

this is voiceover I'm sorry voice over

IP or unified Communications

applications so we're going to get into

more advanced features and functionality

and pieces and parts and you know we'll

cross that bridge when we get there but

this is you know kind of the

certification lineup and if you're going

for your ccnp voice you can see the

exams that you're going to need to pass

in order to achieve that and this is

always a moving Target so you know by

the time you see this it may have

changed or been revised and uh you know

so again cisco.com Go slash

certification is something you're going

to want to check out

this exam 642 437 or C Voice or Cisco

Voiceover IP is going to focus on

teaching you about voice gateways call

legs and dial plans you know so these

are all core Gateway Concepts we're

going to go through a basic

implementation of Cisco unified

Communications manager Express or CME

CME is an iOS based PBX and it'll run on

a Cisco router and it will provide for

call control and call routing

capabilities for IP phones more about

CME as we get into it we'll talk about

Cisco unified border element or cube in

its role as a Gateway and really you

know as a session border controller what

can Cube do for you how does it work

we'll talk about voice related qos

techniques and and you know mechanisms

for implementing these types of things

so you know we talked in the beginning

that this is focused on preparing you to

you know interact with those underlying

networks so we're going to talk about

what you need to know and how those

underlying networks need to be

configured to support the unified

Communications

if we look at the exam blueprint for C

voice Cisco wants you to go through

these following

um you know following key items they

want you to be able to describe a dial

plan describe the basic operation of a

voice over IP call Implement unified

Communications manager Express just to

support phone registration using the

command line interface we're going to

describe the components of a voice

Gateway and we'll talk a little bit

about voice Gateway implementation we're

going to talk about Cube or the Cisco

unified border element that's a session

border controller and you're going to be

able to describe and Implement qos using

a diff serve or a dscp methodology so

again this is just kind of more summary

of what we've talked about already but

this is what the exam blueprint

identifies that you need to be able to

do in preparation for passing and you

know taking and passing your C voice

exam

let's talk about strategies for success

and if you've watched my videos before

you'll understand that you know this is

something that doesn't change and I'm

going to talk about it quite a bit I

believe that you're the master of your

own destiny and you can do this you know

in the beginning when you get into

Internet working Technologies and when

you start looking at collaboration and

voice and video Etc it can really be a

lot to kind of swallow all at once you

know we go abroad and we go deep and we

do both at the same time so my best

recommendation to you is first off

believe in yourself if I can do this

stuff anybody can do this stuff and it's

my goal within this video series to help

break down the things that are

complicated because let's face it these

aren't trivial tasks but let's break

down the things that are complicated

into more simple

concise and to the point facts and bits

of information so that you can digest it

and deal with it in a more manageable

manageable way I'm going to do this by

giving you lots and lots of video

content and a lot of that's going to be

in the form of lecture

we're going to do video Labs or

demonstrations if you will where I will

walk you through configuring something

or setting something up or

troubleshooting something or whatever

and I would encourage you to follow

along with these videos we'll talk about

Labs here in a minute

but be prepared to do what I do yourself

in your own environment

I encourage you to do independent Labs

so take my examples change them up a bit

and do it for yourself you know by

customizing these labs and these

demonstrations you're going to create

problems you're going to break things

and by breaking things you're going to

introduce the opportunity to

troubleshoot and truly learn how to

resolve the issues so you got to play

with the stuff you've got to have

Hands-On

and really are you seeing a pattern here

it's repetition repetition repetition

you got to get in there and you got to

do the stuff and it's going to kind of

grow on you once you get to this point

and you've done the labs and you're

starting to really become familiar with

the ins and outs of how we get things

done within the voice world you're going

to go through your final review you're

going to go and you want to participate

in some exam simulations there's lots of

great resources out on the internet for

that so I'm not going to pitch anything

in particular I would encourage you to

stay away from the brain dump sites to

just try to get you to memorize content

you really can't pass a Cisco exam and

you really can't be a good engineer for

that matter by studying brain dumps now

on the other side of the coin it doesn't

hurt for you to become familiar with the

types of of topics that are being seen

in the current exam Cycles so you can't

memorize questions and answers you're

going to fail but if you can become more

familiar with what to expect and then

polish up your skills in those areas

you're just going to have you know an

advantage when taking the test so I

don't encourage brain dumps but I do

encourage doing some research finding

out hey what's happening in the current

exam series what are the things I really

need to focus on and then you know split

and polish your skills accordingly and

finally take the exam you know it's not

a big deal if you don't pass a Cisco

certification exam the first time around

trust me I've failed plenty of them and

I probably will continue to do so

throughout my career exams are just a

measurement they're a benchmark you know

they're gauging where you are

unfortunately the first time you go in

you're probably not where where you

think you are but you said as an

opportunity to identify the areas that

you're still a little weekend

take the chance to immediately go fill

in those holes in your knowledge and

then go take an immediate attempt I am a

huge huge huge advocate and I really

can't stress this enough of taking a

retake of the exam immediately upon

failing it so there's something like a

five-day grace period that Cisco is

going to make your way between failing a

test and taking it again take your next

attempt on day six or you know whatever

don't go longer than like 10 days or

you're going to start losing the edge

so you know get in a rhythm bang the

stuff out and you can do this you will

need access to real equipment and

whether that's rack rental or whether

that's equipment that you have access to

in your own lab or someone else's that's

fine but you're going to need layer

hands on stuff and we'll talk more about

rack rental Etc here in just a moment

um you know in my opinion rack rental

can suffice but it is really best to

build a lab

it's just more flexible that way but I'm

not gonna I'm not gonna dog rock around

it definitely has its place

why do I need a lab well voice

Technologies they're interactive you

know we're not dealing with things that

can simply be simulated on paper we

really need to touch and feel things

studying for voice is not the same as

studying for a data or a routing or a

switching exam you know we're dealing

with interactive Technologies here we

have telephones we have experiences that

take place over these things and it's

important that you can lay your hands on

it set it up and really make it work and

really see what happens you know so

that's kind of my Approach on the thing

in a real world environment means real

problems and basically what I mean by

that is you're going to screw things up

I do all the time

and that's the opportunity for you to

troubleshoot and troubleshooting

something is when the learning happens

when it's broken you're going to figure

out how to fix it and then you're going

to commit that to memory if it just

works the first time around you're never

going to remember how to do it so

Embrace those challenges and when things

don't quite go as planned they really

can help through the study process and

help you learn the technology

so a little bit more on rack rental can

I use rack rental time absolutely just

keep in mind that the environment that

the rack rental provider has set up for

you may not exactly match the labs that

I'm presenting here in this video they

may not exactly meet your needs but

they'll probably get you pretty close so

keep in mind that labs are kind of a

personal thing and everybody's is a

little bit different

um you know so should you use rack

rental you know it's really up to you

um you know I prefer real gear but that

doesn't mean that I'm opposed to rack

rental now let's talk about a lab

topology and this is an example topology

that I'm going to refer to throughout

this video course and really throughout

this whole video series

although I may vary it just a little bit

as necessary this is an example of the

current state of my lab here and it's

always in a state of flux so don't get

you know fixated on oh that's a Cisco

3750 switch instead of a 3560 switch or

don't worry that oh that's a 2811

instead of a 30 you know whatever don't

worry about model numbers quite that

deep understand the fundamental concepts

and really what I'm going to talk about

with the topology is what you see on the

screen here this is similar to my lab

yours can and may vary and that's

perfectly fine that's okay don't worry

about making things like that understand

what we're trying to demonstrate

understand what the necessary components

are and mock it up accordingly so you

know this is not an eBay shopping list

although these are some fairly good

choices of equipment to get your hands

on if you'd like to use it as such but

this is a good reference example of what

we're going to use here through this

video series

really kind of sums it up right now and

you know that's the intro you know video

for the course here we're going to start

you know with some of the fundamental

concepts and you know talk about

gateways and talk about you know Cisco

UC in general and then we'll get into

demonstrations of how to do things so

thanks for watching I'm really looking

forward to going through this video

course with you and for those of you who

are returning after successfully

completing your CCNA voice welcome and

thank you and I hope I can live up to

your expectations throughout this track

on the C voice exam thanks guys and I

will talk to you soon have good studying

and I'll see you see you later

[Music]

thank you

welcome to module 2 in this ccnp voice C

voice exam video series in this video

we're going to be talking about the role

of a voice Gateway and how it plays into

your unified Communications environments

when we talk about the traditional pstn

and really to understand the role of a

voice Gateway it's important to

understand pstn fundamentals so I'm

going to take a second here and sketch

out to you what the pstn typically looks

like so if I were to go ahead and say

that this circle represents the public

switch to telephone Network the pstn is

comprised of a bunch of various Central

offices and these Central offices are

interconnected and we call these Central

offices cos they're interconnected and

signaled with something we call ss7 and

we don't have to worry about that too

much from the end user perspective but

that's the the magic that kind of makes

the pstn happen now when you have a

location or business and we'll go ahead

and withdraw a couple of them here on

the screen and we have a PBX we'll go

ahead and show our PBX we have a

connection to

the central office PBX and this

connection to the central office is

carried over what we call a local Loop

and typically this is you know an ISDN

PRI this could be a pots line this could

be a BRI it could be a SIP trunk you

know in modern times anyway but you know

we've got these local loops go ahead and

use use ISDN or T1 as the example here

and you know when your call is placed

from your PBX you know in fact on the

back of this PBX you know we've got our

traditional handsets hanging off of the

end of it so when a call was placed from

one handset to another you know so let's

say you know handset a here wants to

call handset B the call is going to go

through the PBX

out to the pstn across these various Co

interconnections and these Co trunks as

they call them and ultimately work its

way to our site to the PBX and ring the

phone that we've dialed now there's a

concept of interconnecting traditional

PB access and we called those tie lines

or Thai Trunks and really what it is is

just another circuit going into ACO

we'll draw a couple different ones here

I'll just call it a tie line

and really think of this as for example

you know point to point T1 where PBX

site a

is connected to the PBX at site B so

they act as though they're within the

same premise or or certainly anyway

they're reachable from premise to

premise but all that signaling is going

through the pstn now things are a little

bit different with voice gateways and

really you know I'll go ahead and I'll

just draw another site here that is a

unified Communications enabled location

and we're still going to have that Co

trunk and it still might be an ISDN PRI

however it's going to enter our premise

and terminate to a device called a voice

Gateway now the voice Gateway is going

to be on our Lan and we're going to have

other things on our land such as our

call manager such as our IP phones Etc

so it's a little bit different

architecture wise than what we've done

traditionally but the same kind of

Concepts apply you know relative to the

pstn

when we talk about call signaling voice

gateways are signaled or have call

signaling occurring with one of a number

of different protocols the four

protocols that you're going to primarily

deal with are shown here on the screen

we have sccp which is the skinny call

control protocol we have mgcp the media

Gateway control protocol we have the

standard h.323 and we have the other

standard sip so lots of choices here

it's one of those things that you'll

have to kind of analyze what is my use

case and ultimately you'll make a

decision of of in a given instance am I

using skinnier am I using sip or you

know what makes sense for this

implementation or this client so these

are the four common call signaling

protocols that you're going to run into

and here in the next couple of slides

we're going to talk about what each of

these call signaling protocols has in

common and then you know not so in

common with the others so that'll kind

of be the basis for understanding how

you make those choices

foreign

CCP the skinny call control protocol

skinny is a Cisco proprietary call

control protocol and it's primarily used

as a protocol for signaling events

between unified communication endpoints

like IP phones and a a PBX you know

Cisco unified Communications manager or

co-manager Express what if you're a CME

on a router it is a client server

protocol and really the one the thing I

want you to take away from skinny you

know at the high level here is that

because it's a client server protocol

it's not real intelligent you know most

of the intelligence is at the head end

the device is just kind of sending

messages back to that call manager and

call manager is doing the intelligent

decision making you know without the

call manager the phone running skinny

doesn't have the capability to set up

and tear down calls on its own so when I

pick up the handset and I go off hook

it's going to send an event message over

to call manager when I press keys in the

keypad it's going to send messages to

call manager when it's time for my phone

to ring call manager is going to Signal

my phone hey play the wave file and send

your ringer so you know all of these

different things because it's client

server you know really the head of the

snake is the the call agent or the call

processing system like call manager call

manager Express so that's skinny the

skinny call control protocol and once

again most of the time you're going to

see scanning in use for controlling

telephone endpoints however you can also

control things such as fxs ports you

know the analog ports so we can connect

station devices to on a voice Gateway

and more about those ports in that

interconnection method later

the next protocol we want to talk about

is mgcp or the media Gateway control

protocol mgcp is standards based a lot

of people seem to think that mgcp is a

Cisco proprietary protocol that's not

entirely true in fact it's not true

Cisco was a key player and in fact

probably the key player in the

development of mgcp but they didn't do

it alone

mgcp is defined currently anyway in the

latest iteration of the standards in RFC

3435 you know previously you could look

at like RFC 2705 and get information

about mgcp as well but mgcp really in a

nutshell it's the successor to the sgcp

the simple Gateway control protocol and

with mgcp the Gateway devices that

you're communicating with using mgcp

they're they're controlled by what we

call a call agent and that would be your

call manager your unified Communications

manager or your unified Communications

manager express your CMA mccp is also a

client server protocol so there's a lot

of chatting going on between the

endpoints and the call agent one of the

major benefits for mgcp and one of the

reasons that a lot of people have kind

of fallen in love with it is that it

offers you the opportunity for

centralized Gateway and what I mean by

that a centralized dial plan

Administration so if I'm in an

Enterprise and I've got you know 27 28

gateways spread throughout my

organization and I'm running mgcp all of

that dial plan logic I mean every little

last bit of it is managed from the call

manager it's not something that I have

to manage individually from Gateway to

Gateway now I'm not saying that that's

always the best you know the best thing

but it is attractive to some people

another interesting capability and

really I don't even call it a capability

one characteristic or behavior if you

will of mgcp is that it performs

something we call q.931

layer 3 back haul so let's say I've got

a PRI and I've connected that PRI to a

voice Gateway and that voice Gateway is

running mgcp well on let's say it was an

h323 Gateway which we'll talk about in

another slide you know that D channel

that signaling channel on that PRI would

be under the control of the Gateway it's

terminated on I.E The Voice Gateway the

router

with mgcp however that q931 signaling is

tunneled all the way or back called all

the way back to the call agent back to

the call manager back to the call

manager Express so it's a little bit

different

um you know it's got some pros and cons

you know we'll learn more about mgcp as

we go but you know it's kind of

interesting to understand how that

q931pri backhaul actually works

h.323 this is kind of the good old

standby it's been around for a really

really long time h323 is standards based

by the itu-t and it's probably if not

the most it's certainly one of the most

widely deployed protocols for voice over

IP type of applications and we're

talking you know well beyond just Cisco

here h323 you know again it's been

around a long time it includes a couple

of other protocols in the family we've

got h.225 call control we've got htt5

what we call Raz which is registration

admission and Status we've got h.245

control signaling and this is also a

peer-to-peer protocol and when I talk

about a peer-to-peer protocol as opposed

to a client server protocol if you

remember when I was talking about mgc

pure skinny those are client server

protocols and really the endpoint is

pretty dumb there's not a lot of

intelligence in the endpoint if you want

to um if you want to really do anything

you get talk to the call manager well

with h323 that's not the case being that

it's a peer-to-peer protocol the h323

device the Gateway in this in this case

can control or can have control over its

own dial plan you know all by itself you

know you can create call rounding rules

within the h323 Gateway that have

nothing to do with the dial planning

call manager and it's kind of different

in that way and you know the next

protocol we're going to talk about sip

has a lot of similarities to h323 as

you're going to see in the next slide

sip the session initiation protocol sip

is also standards based just like h323

is however whereas h323 was a standard

created by the itu-t

um sip is an ietf standard and you know

I'm going to just call it a standard you

know you'll get people arguing with me

about draft standard status and all this

stuff I'm you know it's it's so you know

changing so fast I'm just not even going

to go there but sip is a standards based

protocol it's also widely deployed

um you know sip is everywhere when we

start talking multi-vendor stuff

especially you know sip is kind of

becoming the de factor or de facto go-to

method of integrating system a and

system B

sip uses the concept of a user agent or

a sip UA with SIP user agents initiate

sessions and we'll get into all the user

agent stuff later in the video series

but I just want to kind of give you a

high level here one of the cool things

about sip is that the signaling

Communications is formatted as ASCII

human readable text messages and that

goes to my next bullet here is that we

have the capability to do plain text

debugging I don't have to sit here you

know run a debug capture tons of data

throw it into a text editor and then

take my highlighter out and go find the

pieces of data that are important to me

I can simply sit there and read it off

the screen and see what's going on it's

very intuitive to troubleshoot and

that's one of the things that I

certainly like with the Sip protocol and

sip just like ht23 is also a

peer-to-peer protocol so you know lots

of call routing intelligence native on

the Sip Gateway and sip some of those

protocols I'm talking about it in kind

of a Gateway context here but it can be

used for so many more things sip can be

used as a phone signaling protocol and

in fact when you do that the phone

itself has the capability or the

intelligence to be able to do call setup

and tear down on its own so I'm kind of

digging deeper than we need to right now

but sip is the last protocol that I

really wanted to talk to you about in

this introduction to the role of voice

gateways so this kind of wraps it up for

this quick little video I just wanted to

talk about gateways what they are what

they do and you know what protocols are

at play we're going to have videos

coming up next and we're going to go

over some of the specific

um you know Hardware models of voice

gateways we'll talk about the kinds of

modules that will go in them and and

you'll get a whole lot more

understanding about how voice gateways

work and ultimately we'll go through

some some pretty significant

configuration exercises so for now this

is your primer on voice gateways and the

role of voice gateways and I want to

thank you for watching and I will see

you in the next video good studying

[Music]

thank you

[Music]

welcome to module three and we're going

to be discussing UC deployment models

and more specifically or I should say

for the most part this conversation

centers around Cisco unified

Communications manager we are going to

mention call manager Express and even

third-party pbx's in this a little bit

but you know by and large this is a call

manager conversation and again this is

part of the C voice exam series uh right

now the exam is 642 437

but uh let's uh let's jump right into

this and talk about UC deployment models

now there are three models that you're

going to run into in fact two two out of

the three you know really most of the

time the third is kind of an exception

to the rule but when deploying call

manager you've got the single site

deployment model

the multi-site centralized call

processing deployment model and the

multi-site distributed call processing

deployment model single site works great

scales well some of the multi-site stuff

also can can kind of incredibly scale if

you really need to get into these huge

you know 100 000 phone count kind of

things but uh you know we'll jump right

into that so keep in mind there's three

single site multi-site centralized call

processing and multi-site with

distributed call processing

let's talk about the single site

deployment model so when we're talking

about a single site we're talking about

one building

local area network only no Wan and we're

talking about a single cucm cluster now

a single cucm cluster from a test

perspective I want you to think the

number thirty thousand but understand

that this has grown you know in the 9.x

version now it's forty thousand sip or

skinny endpoints so this is IP phones

we can support up to 2100 h323 devices

so those could be h323 endpoints or you

know gateways running the h323 protocol

and everything that leaves a site is

going to use the psdn because there is

no connectivity to any other sites

there's no way in there are no other

sites as far as that's concerned so

everything's going to the pstn

some of the benefits of the single site

deployment model obviously it's simple

so ease of deployment you know you've

got one call manager cluster phones

gateways you're done

we've got a common infrastructure for

our voice and our data traffic so and

I'm kind of mentioning you know routing

and switching here you know we're using

that same Enterprise infrastructure

because everything's going to the pstn

and then not to other sites we've got a

simplified dial plan and again because

there are no other sites

we're using one codec so we don't

require transcoding

what are some of the design

considerations for a single site and

really this is what are some of the

design considerations for our call

manager cluster you know this will scale

to you know multi-site as well you need

to provide a fault tolerant underlying

Network infrastructure so I'm talking

routing I'm talking switching here you

need to analyze your dial plan and your

calling needs you know how many digits

do I want to have in my extensions you

know what are my gateways and pstn paths

we're going to talk about codec use and

with a single site you're typically

going to be using a high bit rate codec

such as g711 or more recently g722

we're going to need to evaluate High

availability requirements for the

telephony environment do I need multiple

call processing nodes you know do I need

subscribers if so how many do I need

multiple voice gateways if so how many

do I need multiple paths to the psdn IE

circuits if so how many

so all of these things are going to come

into play you know even with a single

site and then obviously we'll think

about them in multi-site environments as

well

when we look at the multi-site

centralized model we've got a single

call manager cluster still this is

really just building on that single site

environment and we've got you know

support for the same number of endpoints

you know 30 to 40 000 endpoints 2100

h323 devices but here's where it starts

to differ we're going to talk about a

maximum of 100 locations we're going to

start using multiple codecs because

obviously our Wan links are going to be

lower bandwidth than the Lan so we're

going to need things like transcoding

services and we'll use dsps to do that

we'll talk about redundancy options

relative to site survivability and what

I'm going to call srst you know how do I

deal with wind failures and I'll draw a

sketch here in a minute to demonstrate

some of these things

we're going to have calls going across

the land so obviously they'll be cost

savings you know why are calls going

across the land because they can you

know why would I send a call from Phone

8 to phone B in two different locations

across the psdn and pay for it if I can

send it across my data Network for free

and the same goes you know with toll

bypass if I can send a long distance

call

um you know out a Gateway so it's not a

long distance call anymore I bet that

call is going to be cheaper so let's

leverage the land to do that now when we

talk about these multi-site models and

in fact the centralized call processing

model let's look at it this way let's

say that I've got you know a

headquarters location HQ

and I've got you know multiple nodes in

my call manager cluster and phones are

registered Etc you know I may have a

voice Gateway with pstn

connectivity you know let's draw a

little Lan here and I may have a Wan

router you know out to some kind of

ipuan this could be point-to-point links

this could be an mpls Cloud this could

be dmvpn this could be any number of

different things and you know I'm going

to have you know a branch location we'll

call it branch

a and we'll have you know a lan and you

know some kind of router or terminating

equipment for you know whatever that

connectivity is you know and when I have

IP phones out here now I'm not going to

have a call manager cluster centralized

call processing means everything happens

in a central location so in this example

it's happening out here at HQ

now when we look at the branch you know

I mentioned land failure let's say this

ip1 goes down what am I going to do well

this Gateway right here is going to need

the ability or need the capability to

handle local call processing so this

phone can register to the Gateway which

has you know it's pstn pathway you know

maybe it's a PRI maybe it's a plots line

whatever but you know in the event of

that Wan failure the phones can swing

over to the Gateway and the Gateway can

provide for services until the call

manager is is restored

let's talk about the multi-site

centralized deployment benefits

um you know it's just like having a

single cluster but we're serving

multiple sites I'm sorry it's just like

I have a single site but we're serving

multiple sites it's still a single

cluster we're going to have reduced

infrastructure over what would be

required if we had a cluster in each

location which is what we're going to

talk about in the multi-site distributed

architecture we're going to have a

centralized dial plan you know because

there's one cluster all the dial plane

logic is all in one place and we're

going to start having you know

survivability capabilities for these

remote locations should that the primary

call processing node or cluster become

unreachable so that's srst

what are some of the design

considerations for the multi-site

centralized environment obviously we've

got to consider Wan latency and

bandwidth and quality of service and

fault tolerance of the Lan you know we

want to make sure that that Wan you know

is is less than 80 milliseconds round

trip or you know we're going to have

some issues we're going to think about

using different codecs for calls across

the land versus locally you know if I've

got a switched environment and I've got

bandwidth to spare or you know unlimited

bandwidth if you even want to think

about it that way at my disposal the

choice of codec that I'm going to use

isn't going to be such a big deal but if

I'm going to go across a low bandwidth

we in link like maybe a fractional T1

I'm going to start to care about those

things so we'll talk about using low bit

rate codecs again we're going to

evaluate our survivability needs for

srst do these phones need to be able to

register and have pstn access in the

event of a way of failure

and we're going to start thinking about

calling mission control again this is a

man with discussion you know when we

talk about CAC let's say my Wan is big

enough to support 10 calls period

if that 11th call is allowed to happen

guess what happens to all 11 calls they

all start to suck because every one of

them they're all going to be competing

for bandwidth and there's not enough

bandwidth for all of them so qos can't

solve this problem I can't prioritize

traffic when the bandwidth exceeds or

the bandwidth demand exceeds the

available bandwidth so call emission

control can help to Route calls you

either reject calls or you know using

alternate routing you know AAR stuff we

can send it out to the pstn and Route

Around the land so we'll talk about that

stuff in other videos but uh you know

we'll start to think about CAC in the

multi-site centralized design

when we start scaling that even farther

and going into this distributed call

processing model for multi-site we're

going to deal with multiple cucm

clusters and let me draw you a picture

here so you know HQ right here you know

maybe this is a large Branch or maybe

this is some kind of a campus

environment you know maybe it's college

and I have you know Regional locations

or whatever you know and I've got this

big monster ipan in the middle

and I've got this call manager cluster

you know out here and he handles his

local phones and I've got this call

manager cluster out here and he handles

his local phone so that's a multi-site

distributed call processing environment

again you know thirty thousand or forty

thousand employees per cluster

the same kind of counts on h323 devices

as you've seen before but as we start

doing this multi-site environment we're

going to evaluate the use of things like

Gatekeepers and deal with ways of

Distributing dial plan

we're also going to continue to think

about things like pstn toll bypass so

lots of stuff you know come into play

here what are some of the benefits the

multi-site distributed environments

obviously we're still using the IPO and

for Enterprise calls so we're not

sending things to the pstn

um you know unless they need to go to

the psdn we've got tool bypass

capabilities still we're maximizing your

bandwidth utilization still you know

because we've got voice and data running

one common Network we're gonna you know

have benefits of you know High

availability and call processing

survivability we're gonna have lots and

lots of scalability because obviously if

I can get 30 40 000 endpoints per

cluster I can put the things all over

the place I don't care if there's 10 or

100 locations or even more you just keep

spinning these things up and keep having

you know more and more capabilities what

are some of the design considerations

for the multi-site distributed model

well you're going to start thinking

about and I mentioned it before but

you're going to start thinking about

things like Gatekeepers and sip proxies

to deal with Distributing your dial plan

you're gonna have to think about trunks

between systems obviously if you you've

got multiple clusters you've got to have

ways for calls to flow between those

clusters those are going to be sip club

I'm sorry sip trunks or intercluster

trunks

you're going to want to think about

things like using hsrp for gatekeeper

pairs to give you some redundancy and

access to those Gatekeepers you're going

to be limiting the way into one type of

audio codec again this is Gateway

gatekeeper stuff and we'll get into that

a little bit more as we go on but lots

of things go into these multi-site

distributed environments

when we start talking about these

distributed environments and really this

is more a

it's kind of a blend it's it's not a

single site call processing model it's

not really distributed but let me just

draw you a picture let's say we're in an

Enterprise and we have two data centers

and these two data centers are kind of

the center of our universe you know dc1

and dc2 I may have a call manager Pub

and sub

in this primary data center and I've got

this ipwan out here and I may have

another sub at data center 2 and you

know maybe all my branch locations are

connected you know let's call this a

branch router you know maybe they're all

connected via you know multiple Wan

links one to each data center you know

and I've got IP whoops IP phones out

here you know like that

Etc so I've got to think about you know

what's the delay across this land you

know because I've got nodes here that

need to be synchronized with each other

so what is the delay of this win and

like I'm showing you here it needs to be

less than 80 milliseconds you start

creeping above those numbers you're

going to have problems we're going to

start thinking about qos enabled way and

links and really we're not just going to

start thinking about it these things are

going to be critical we need to make

sure this communication is happening in

a timely fashion

we need to make sure that we're

minimizing Jitter we need to make sure

we have sufficient bandwidth and we need

to manage packet loss and errors if this

Wan is lossy and drop in 10 of the data

you can bet that you're going to notice

it and I mean that's just all there is

fit you're going to notice it it's going

to be a problem so you really have to

handle your network and make sure things

are in order to do this clustering

across the land but this is kind of a

hybrid it's it's multi-site and it's

still kind of distributed call

processing but it's still just one

cluster so cool stuff here I'm seeing a

lot of this so clustering across the

land definitely something you're going

to run into out there in the Enterprise

you know when we talk about these you

know multi-site environments especially

the distributed call processing

environments we may not just be talking

about call manager here we may be

talking about cucm you know call manager

in one location we may be talking about

CME or a small location somewhere else

with a SIP trunk to call manager we may

be talking about a traditional or Legacy

PBX with a voice Gateway running sipta

bring it into the thing and you know

these things could could look like lots

of different things you know I could

have you know HQ out here

you know he could have call manager and

you know so CM we could have another

location out here with a CME running on

an iOS device we could have you know a

voice Gateway at another Branch

connecting to a traditional PBX

so lots of different models here in lots

of ways these things can all connect

together

I know you've been drinking from the

fire hose through this video and uh I

both apologize for that and also say

you're welcome because you know this is

really kind of where the rubber meets

the road here on call manager design you

need to understand the multi-site you

know distributed and centralized model

you need to understand the single site

uh model and you know how all these

things kind of tied together so I'm

going to call that good for this video

and in the next video we're going to

start talking about some of the specific

hardware and jump kind of back into the

gateway conversation and talk about

makes and models of Hardware types of

interfaces and and what's typical what

are things you may see out there in the

field what are the things you're going

to need to know for the exam and you

know really continue the Gateway

conversation so thanks for watching I

really appreciate it I hope that this

has been informative to you and good

studying I'll see you in the next video

[Music]

thank you

[Music]

welcome to module four this is going to

be a really quick review or overview of

common voice Gateway Hardware that you

may run into in unified Communications

environments we're going to talk about

the modern stuff as well as some of the

more traditional pieces of equipment

that you're likely to run into in the

field

pstn gateways you know we've got all

kinds of gear here that we can be

talking about you know lots of routers

lots of you know purpose-built devices

Etc but let's talk about the more common

things you know when I show over here on

the left modern equipment we're talking

about generation 2 Integrated Service

router so ISR G2 Hardware so you know

the X9 series of things so the 2900s the

3900s ETC one of the cool things about

this uh you know the router based voice

Gateway services

is you can provision the hardware

however you want you know you can put

fxo modules or fxs modules in it or you

can put a an e m module or a T1 PRI

module or you know if you need to do you

know something other than that you know

you've got choices here as far as

quantities of ports and densities and

exact configurations so ISR G2 very

popular choice here

you know and you can scale this from a

single pots line clear up to multiple

pris uh taking a step back you know one

generation we're going to be looking at

the original integrated Services router

so your 2800s your 3800s Etc very

similar in functionality

to the more modern you know ISR G2

Hardware uh many similar cards you know

we're going to have the same type of

flexibility and configuration where

we're dealing with you know fxos and

fxs's and E Ms and PRI Etc so these are

the two you know you know tried and true

um you know very popular widely deployed

types of equipment you're going to deal

with you know the isrg ones g2s I've

done a lot of work with gateways even

older than this you know Cisco 3700s

um you know if you even want to go back

farther

um you know the 3600s you know 3660 made

it great voice Gateway with you know ISD

and PRI or T1 cast circuits in it and

you know same goes for you know an old

device we called the um the vg200 you

know it was a very small 1u voice

Gateway that was used in a lot of

entry-level deployments or you know

small density deployments but let's move

on to the next slide and we're going to

talk about some of the special purpose

gateways that you're going to run into

these are more purpose-built devices the

the ATA 187s or 88186s those are the

analog telephone adapters now the ATA

whether you're talking the original 186

is 188s or the more modern replacement

the 187 you know they serve generally

one purpose and that's to provide an

analog station connection to your ippdx

a lot of times you'll see these

servicing fax machines or servicing

um you know an analog phone on the wall

for you know maybe it's in a warehouse

and it's just a common area phone things

along those lines where you know perhaps

the distance limitation of ethernet is

an issue but you can still swing a pots

line out there so analog telephone

adapter is the biggest difference

between the ATA 187 and the more Legacy

186 is 188s the ATA 187 is sip based and

the 186s were skinny so a little bit

different

um but you know serve a similar purpose

if you want to kind of kick it up a

notch to an ATA on steroids or you know

and really it's a small miniature voice

Gateway it'd be something like the vg202

or 204 you know get a couple of ports of

fxs uh connectivity now these are iOS

based devices whereas the smaller you

know ATA 186 187s you know those are

kind of purpose-built devices and they

really don't have you know an iOS like

interface that you're used to on a

router but you know you get into the

vg202s ETC and they're going to be

another voice Gateway for high density

analog connections was the vg224 or more

historically the vg248 now it might

sound like the 248 it's bigger well it

was bigger but it was older you know the

248 was a Cisco acquisition you know

early you know first second gen product

whereas the vg224 is a you know an iOS

based platform so two very you know

different pieces of gear the v248 you

know that could run skinny the vg224s

you know those are going to support

multiple protocols you can you know mgcp

sip h223 you know we can keep going on

but uh those are your analog Services

gateways now you can still put a couple

of analog ports or you know a lot of

analog ports if you want into one of

those is you know ISR g2s or ISR g1s but

it's not real common

um you know so for dancing you're gonna

end up in something like a vg224 more

often than not these days now Universal

gateways you know we just call them

Universal gateways because they can be

configured to do lots of other things

and really so could the isrs but these

are more service provider-centric or

more specific purpose-built devices you

know the as5350s or as5400xms

um you know those were you know I'll

call them carrier grade equipment you

know they're built a little bit

differently than gear that you're

targeting for an Enterprise

um you know Cisco 7200s you know

especially things like 7206 vxrs those

were not real popular you know yeah I

I've maybe seen like one maybe in the

field but uh you know people just didn't

use them for voice you know excellent

wind routers excellent Edge devices just

not real popular for voice but uh you

know certainly there were some

capabilities there

um you know as we get into talking about

you know different routers and

programming things you're gonna likely

kind of adopt a favorite device that you

use you know mine is the ISR you know

ISR G1 and G2 those things are just

amazing you know you can provision them

however you want but this was really

meant to just be a quick tease of what's

out there and give you an idea of the

types of Hardware you're going to run

into you know there's not a whole lot

that we're uh we're going to dig into

great detail on there but uh you know

one more thing I want to hit in this

video session border controllers it's

another type of Gateway and you're going

to hear me refer to these things as

cubes or Cisco unified border elements

more often than not but a session border

controller is a type of Gateway that

really I want you to think of it as an

IP to 8p Gateway in fact early on that's

what Cisco called it they called it ipip

Gateway as a feature set in one of the

uh you know the the iOS feature packs so

session border controller is used to

terminate media sessions and signaling

and there's different ways to configure

these things with media flow around flow

through and we're not going to get into

all this in this video but uh they're

often installed as an edge device so

let's say I want to have a SIP trunk to

my carrier I'll have them bring in you

know an IP connection to my Edge and

they'll plug it into my queue or my

border element my session border

controller and then I'll have a SIP

trunk from that device over to my call

manager and you know it sits there at

the edge of the network

um you know providing that point of

demarcation

um these things are providing for sip to

sip h323 the Sip or h323 to h323

connectivity so again it's an IP to IP

Gateway and one of the biggest

differences is when you look at a

traditional voice Gateway

and we'll cover a lot of this when we

start talking about dial peers and

programming those things but you're

going to have a a pstn lag or psdn

interface

and Associated dial pair what we call a

pot style pair and you're going to have

a IP facing interface um you know

whether that's a SIP trunk or h323 or

whatever

and you're going to have you know the

associated VoIP dial pairs but the

session border controllers it's IP on

both sides so sip to sip h33 to sip Etc

so that's really it for talking about

you know types of Hardware gateways that

we're going to get into

um you know really you're going to be

touching and feeling these things and

getting your arms around them so there's

not you know a real need to go a lot

deeper than that but I just wanted to

give you a quick little nickel store of

the types of Hardware that's out there

which you're likely going to run into in

both new and existing installations so

thanks for watching guys as we get into

the next series of videos we're going to

be talking about voice call legs and

going through basic dial pure

configuration and some not so basic dial

pure configuration and really start

talking about how to make the calls flow

through these devices so thanks for

watching uh good luck with your studying

and I will see you soon

[Music]

thank you

in this video we're going to talk about

call legs and dial pure fundamentals I

want you to understand the concept of

how calls route their way through a

voice Gateway and that's all about dial

pairs so let's just jump straight into

this I think that a picture is worth a

thousand words as far as this subject

goes and I think we're going to help

paint you that picture as we go through

through this video

now when you're dealing with the voice

Gateway I want you to understand that

every call into and out of a voice

Gateway requires a dial peer to be

routed

and we're going to make decisions on

routing based on calls coming into us or

what we call an incoming call lag we'll

go ahead and draw you on here

and calls going out of us and what we're

going to call an outgoing call leg now

you know inbound outbound Etc so a voice

Gateway is going to make decisions that

will influence how it routes a call when

the call enters it

through the incoming Diop here and as

the call leaves it

on the outgoing dial pair or call leg

and really call legs and dial pairs are

not synonymous to each other but they go

hand in hand from a functionality

perspective so if you've got multiple

hops in your voice Network where you're

jumping from voice gateway to voice

gateway to voice Gateway perhaps you've

got sip trunks or h.323 trunks between

them you're going to have dial pairs at

each point and you're going to have

colleagues at each point so let's just

jump right into this here and talk about

types of dial pairs there are three

types of dial pairs I want you to be

concerned with pot style appears VoIP

dial piers and multimedia over IP dial

pairs in fact the first two you're going

to use day in and day out and on

occasion you may use an mmoip Diop here

so first and foremost pot style appears

pot style piers are used whenever we are

going to route a call in or out

a physical interface that is either

analog or TDM so fxo and fxs ports T1

Cas interface or an ISDN PRI interface

VoIP diopiers we're going to use these

when we're routing a call to another IP

peer so that could be router to router

you know it could be an h323 connection

to another voice Gateway it could be a

SIP trunk to a call manager Etc

multimedia over ipdial Piers we're going

to use when we're doing store and

forward facts and they are going to

point to an email destination and more

about these types as we go through the

various videos in this course but I just

want to jump right into it and we're

going to show you what dial Piers look

like and how they work so keep in mind

in fact this is a good analogy to build

is that together dial peers build a dial

plan so if you want to think of this

similar to routing I want you to say

routes

build a routing

table

so dial piers are really similar to

Ralph's and the dial plan is really very

similar to the routing table now I've

got a voice Gateway here in the lab that

I'm going to go ahead and show you

some dial up here is on this is a Cisco

router with a analog fxo module in it

actually it's a two Port fxo module and

I use this as a connection to a pstn

interface for the lab here I happen to

use nine as an outside access code so

whenever I pick up my IP phone I press a

9 and you know I dial the numbers I want

to dial well what happens in my call

manager is I send the calls to the voice

Gateway and once I hit the voice Gateway

it's going to evaluate its dial Pairs

and look at an incoming called leg

and what it's going to do is it's

actually going to match based on some

values in the Gateway

for an incoming call leg and make some

rounding decisions and then it's going

to look at what number I'm trying to

dial and try to find an egress path and

it's going to match another diode here

and I want to show you here on the on

the inbound dial up here this is coming

from you know call manager so I've got

an iPhone with an IP connection to call

manager so the incoming call leg is

actually going to match this dioper

voice 100 VoIP and it's just going to

match it I'm not doing anything with it

you know but it's going to know that it

came from this you know the source

address Etc I've got my income in

figured out I need to pick a Next Top I

need to go to the pstn you know my call

is nine let's say I dialed a toll-free

number you know maybe I dialed you know

like a nine whoops let me draw it here

nine one you know five one three five

five five one thousand or whatever

um

you know maybe that's the number I

dialed my router or my voice Gateway is

going to need to analyze the number I

dialed and figure out where to send me

now this dial pure voice 12 pots that

you see here it says destination pattern

nine one and then digits two through

nine and then you know X number of more

digits so basically I'm going to match

that pattern like I would a route so I'm

going to match that dial up here and

then it says Port one zero zero well

that happens to be an fxo Port where

I've got a pot sign connected and as

simple as that the call is made you know

the routing takes place and you know I'm

talking on the phone so dial peers

really don't have to be crazy

complicated although they can be there's

lots and lots of things you can do with

dial pairs tons of different ways you

can match on things coming in and out

and there's orders of priority and what

do I look at first and we're going to

get into all of those things but at a

high level I want you to understand dial

peers are like routing statements we're

going to match one as we come into a

device we're going to match one as we

leave a device and you know really high

level that's it you've got it so as we

go into more explanation of dial pairs

just keep that in mind there's incoming

colleagues and outgoing call legs each

call leg is going to be associated with

a incoming dial pair and an outgoing

Diop here with that we're going to say

thanks for watching and we're going to

move into the next set of your videos

and we're going to really dive into dial

pure configuration in a lot greater

detail this was really just a skim on

the surface so I'll see you in next

video good studying and we'll talk to

you soon thanks guys

[Music]

in this video we're going to talk about

dial pair matching and call routing

decisions as they're made within the

voice Gateway now previously we talked

about the structure of dial piers and

how they relate to you know call legs in

fact here's a review of the call leg

concept you know where I've got a voice

Gateway and I've got an incoming

call leg and then I've got an outgoing

call leg or an outbound call leg

and we talked about how we would match a

dial pure for the incoming leg and we'd

match a dial beer for the outgoing leg

and that's how we decide how to make a

rounding decisions

and we talked about how together

all of these dial Piers like routing

statements would create a dial plan or

you know effectively a routing table so

that's all the same and we're going to

kind of zoom in on it a little bit

granular a little more granular detail

in this video now I told some fibs about

how things were actually working in the

last video for the purpose of keeping

things simple I didn't want you to focus

on the exact specifics of why I'd match

this pier that Pier you know the details

of what came first or any of that I

wanted you to focus simply on the fact

of understanding incoming call legs or

inbound colleagues you know whatever

term you want to use an outbound or

outgoing call legs and I wanted you to

understand that we would evaluate dial

peers and match them to make routing

decisions beyond that you know the rest

of the is for this video and you know

stuff that's going to follow

I want to go a little deeper and I want

to talk about the specifics of what we

match why we match how we match the

order in which we match and all the

various considerations that come into

play with that so let's now zoom in on

dial Pairs and talk about you know the

meat and potatoes of it if you will and

I've got a good example for you here of

pots dial Piers this is a fictitious

Network here we're gonna you know go

ahead and put some extensions 1001 and

2001 on these phones and I'm going to

walk you through

some dial Piers that would be on The

Voice gateways shown here on the page

so let's say extension 1001 wants to

call extension 2001 so that's the call

we're going to create well what's going

to happen is extension 1001 is hooked up

to an fxs port on the router and let's

assume

that the router is connected to the pstn

Via a PRI

and that the router at the other end is

also connected via a PRI and we've got

another fxs Port so let's walk through

the flow so guide extension 1001 picks

up the handset

gets a dial tone and makes a call he

dials two zero zero one so here's what's

going to happen and obviously these

patterns wouldn't work on the real pstn

but we're just mocking this up as an

example here

so

we're going to match an inbound dial

pair well I want to show you dial pure

voice 20 pots down here I've got a

command here incoming called dot that

command is going to match

any dialed number

coming in so I'm going to see that in

the dial period I'm going to match that

so I'm going to match an incoming dial

up here so I've matched 20.

I have

now that I've matched an incoming diab

here I have the ability to evaluate the

dial plan think of it as a routing table

and look for my next top well I've

dialed 1001.

well guess what I'm going to use the

same dial beers and outman dial beer in

this example maybe I shouldn't have done

that for the example but it'll work just

fine because my destination pattern is

1001.

so I'm sorry I'm lying to you

let me back up for a second

my destination pattern is actually 2001

I'm going to match this style I was

looking at my drawing backwards so I'm

going to match 20.

for an inbound dial pair I'm going to

match 10 for an outbound dial pair

so I've matched 10 for a destination

pattern of 2001. it has Port 200

configured that's this interface right

here

so we're going to send the call at that

port

to the pstn

so we're gonna have

call come in you know the magic of the

psdn is going to Route it to me and

we're going to match you know another

incoming call leg now if you'll see dial

pure voice 10 pots again incoming call

Dot that's a beautiful match for an

incoming dial pair you know I'm going to

match the dial number

so now that I've matched an inbound dial

up here number 10 let's match an

outbound dial here well you know that I

dialed 2001

so since I dialed 2001.

again this command down here destination

pattern 2001 port one zero zero we're

going to send the call out Port 100 and

we're going to ring the phone

the extension 2001 is going to answer

and our call is connected so we've

actually in this scenario evaluated four

dial pairs we had the inbound call Lake

to the Gateway the outbound call leg

from the gateway to the pstn

the inbound call leg from the psdn to

the Gateway and the outbound call leg

from the PS from the gateway to

extension 2001. so we've walked you

through the whole cycle and that's how

the call was made I've thrown a lot of

different options at you here you know

I've introduced you to incoming called

Dot

we'll talk more later about this direct

name or dial command that has to do with

you know one stage versus two-stage

dialing but really what I want you to

take away from this is again understand

your matching incoming dial Pairs and

your matching out loud dial pairs

understand some of the options under the

Diop here like port

um understand forward digits you know

I'm telling it in this case the four

digits all so all the digits I dialed

um you know continue to forward them

and uh you know really that's a good

textbook example now

let's talk about the matching because

I've showed you that we matched but I

didn't really explain why there are a

few things you can match on

first Dennis the dialed number

information service

that's the number that was called

and you saw I used incoming called

number as the command

I can match on Annie which is my caller

ID using answer address

I can match on Annie using destination

pattern

I can match based on The Voice port

and if I don't match anything else I'm

going to match dial pure 0. and dial

peer 0 is our default dial pair so let's

talk about dial pair zero

dial pr0

is not a dial period program

it exists by default within the iOS

what is dial peer zero support well it's

g729 or g711 only

any packets

that uh match dial up here zero

are going to have an IP precedence of

zero not great

that or voice activity detection is

enabled on dial pure zero

there's no support for the RSVP qos

mechanisms when using dioper 0

and it does not support direct inward

dial and that's a big deal

we'll talk about one stage versus

two-stage dialing throughout the course

let's suffice it to say that if you call

a gateway

and your inbound call leg matches the

default dial here zero

instead of your call routing to the

number you dialed

you know we call that one stage dialing

you'll be greeted with dial tone which

is you know you then have to type in the

number and then it's going to evaluate

the route plan and find a dial up here

and write your call

that's not good you know we used to do

that back before director dial was

available

and we did two-stage dialing so the PBX

would answer give you a second dial tone

you'd dial the destination

but nowadays you know we don't do that

anymore

now we allow the pstn to pass the DNS to

the PBX

and we use one stage dialing

now I want to show you an example of a

real voice Gateway this is Gateway that

I use

here in my lab

but it's not lab gear this is actually

production gear show

dial peer voice summary good command to

use we'll go ahead and blow this up a

little bit bigger

and I'm going to show you let me run

that again that screens a little wire

I'm going to show you the dial peers I

have configured in the Skateway

I've got three pot style pairs a 10 an

11 and a 12.

and I have two VoIP dial Piers a 100 and

a 200.

you can see

the patterns that I'm using

they're a bit more specific than those

vague patterns I was showing you in

previous videos

and this is actually a Gateway I used

for psdn connections

so if I match nine

plus a number

you know the second digit beginning with

two through nine and six more digits so

I'm dialing a seven digit number

then route it out for one zero zero that

happens to be an fxo Port going to an

analog phone line here

the next pattern that's a 10 digit match

and these still aren't perfect patterns

for a production system you know your

patterns are going to be a lot more

specific than this and that's okay I'm

just trying to demonstrate some Basics

here

um for inbound calls that dial pure 100

you see if a call comes in and presents

A genus of you know one through two dot

dot dot you know so 1 000 numbers or two

thousand numbers then it's going to send

it to an ipv4 Target you know that's

going to send it to a call manager

so check this out I'm going to show you

a little more detail of what's running

in these dial pairs Chevron pipe section

dial here

voice

so

I have got

actually here dial pure voice 10 Bots

destination pattern you know nine plus

seven digits

I point it to my analog port

and I say forward digit seven so I'm

telling it to strip the nine because

obviously that's an eight digit pattern

and I only want to match a seven digit

number

the number 11 that's a 10 digit number

yeah so you're seeing four digits ten

you know plus the nine of course

but we don't want to pass the night into

the psdn

you'll see my VoIP guy up here which is

a sip Diop here in this example you know

I can do sip for h323 and within VoIP

Diop here this one happens to be a sip

Diop here

so I'm creating effectively a SIP trunk

to my column manager

you're going to see I've you know made a

codec selection in there g711 EULA

and then I've got you know dial your

voice 12 plots that's a long distance

number a nine plus one plus ten and

that's coming out my analog part you'll

see forward digits 11.

and then you'll see my incoming dial up

here

that's 200.

incoming call number dot now I want to

tell you

you can use the same Diop here or

diopiers

as incoming and outgoing dial pairs

there's no reason I couldn't have gone

to die up your voice 100 VoIP and set

incoming call Dot

and you know not had to have 200.

but I broke it out Simply to make it a

little easier to understand

so that's a good example of a production

Gateway

and uh actually let me show you

something pretty cool here let's do a

debug in fact I may already have it

enabled should debug I do already have

it enable term mon this debug is going

to illustrate dial pure matching

as I make a call

to the pstn so I'm picking up a phone

connected to this Gateway

and I'm going to dial 9.

and you see

called number equals nine I'm going to

say one

six one four

five five five one two one two

so you're actually seeing my dial up

here

evaluation process as it occurs as I try

to decide which diet appears to match

and

I did that slowly so you could see how

the digit matching process works

on

a voice Gateway

digit matching and dial peer analysis

happens

as each digit is dialed so the first

match I find

I will take I will use it

it is not the best match it is not the

most specific match it is the first

match

so keep that in mind and you'll find

that that's the opposite of how call

manager does it in call manager we're

looking for the best match so you know

some differences there

time

we'll close the session here and show

you one more example this is a little

different this is a voice Gateway with a

pstn connection on one side and at PBX

on the other a call manager

so we're going to have again let's

assume PRI and let's assume this is an H

dot three two three connection

call comes in

from the pspn over the PRI

and you know presents a you know a four

digit DNS you know I dialed you know

2001. so what I'm not showing you here

is you know this is a land and out here

off the switch there's a phone you know

with extension 2001

I'm going to match this incoming call

number Dot

and I support direct inward dial so the

Denis that was dialed you know from out

here on the psvn you had a guy with a

phone and he dialed

2001 presumably

routed across pspn

hit my incoming call lag match the style

for your voice one pots

and I support direct number dial so

we're going to support single stage

dialing

next I'm going to evaluate my dial Piers

so the number dial was 2001. ah I see a

match destination pattern 2001.

let's avoid dial pure you'll see it here

and the session Target ten one one five

that's the PBX out here 10.1.1.5

so I've matched the incoming and the

outgoing dial up here and I connect the

call

if the PBX wanted to make a call

he has his own dial plan

that will send the call to me

and uh you know I'll match

you know on an incoming Diop here in

fact I'm going to match the same one one

because incoming call number Dot

I'm then going to evaluate

to make an outbound call now he had a

nine that he dialed as a prefix code for

an external call

and actually this is a really really

good destination pattern to use this is

more of a production quality pattern

nine and then a digit of two through

nine and then two more digits

and then the digit of two through nine

and three more digits so it's a seven

digit number

following the rules of the North

American numbering plan

and this boarding to Port two zero zero

with a forward digits of seven so we're

going to strip that nine so call is

going to go out to the psdn as you know

five five five one two one two

so there you have it there's an example

of a pot style pier and a VoIP Diop here

on a voice Gateway

so we've talked about a lot we've gotten

into a little bit more detail on dial

pure configuration

you've seen some commands show dial pure

voice summary and you know the debug

dial pure stuff that we can do

again you know essential Tools in your

toolbox when dealing with dial peers and

call routing within a voice Gateway

in the next couple of videos we're going

to talk about some of these interfaces a

little more specifically some of the

analog you know fxo and fxs interfaces

and we'll talk about you know some

analog signaling techniques

and uh you know really start introducing

you know things like e m to yet you know

some more specific configuration related

to these ports so again you know we're

starting at the thousand foot View and

we're zooming in a little bit at the

time so with that I'm going to say

thanks for watching and uh good luck

with your study I hope that this has

been informative to you and I'll see you

in the next video thanks guys talk to

you later

[Music]

thank you

welcome to the module on configuring

basic ISDN PRI services

thanks for hanging in there I know we've

done a lot of slide content through the

last you know six or so videos you know

we're hitting a lot of theory

information but we're going to change it

up a little bit in this module and we're

gonna get practical we're going to

configure ISDN PRI on a Cisco router or

voice Gateway and I'm going to walk you

through various configuration steps to

setting this up we're going to actually

place a call from an iPhone on a call

manager and send it out of Gateway and

across the PRI and and bring things into

service I'm going to show you some

useful debugs and uh you know give you

the all-around you know Nichols tour the

ISDN PRI now this is important both from

a practical perspective as a voice

engineer but also

um for the certification exam so Cisco's

going to want you to know this really

really well and you're going to use this

day in and day out so let's talk about

ISDN PRI so integrated Services digital

network is ISDN isdn's been around for

you know decades and primary rate ISDN

is 23 Bearer channels or B channels and

one Delta Channel and we use the B

channels for supporting up to 23

simultaneous calls

um the signaling for those calls occurs

in the Delta Channel and the audio is

going to ride on the bearer channel so

we're talking Standard um you know

Nyquist theorem here you know 8K times

8-bit sample rate equals 64k call you

know it's a a physical T1 that's

important to understand the ISDN PRI is

just a T1 and it's got an ISDN switch

connected at one end and up PBX

connected at the other so physical D1

Services

um 24 time slots you know one for call

signaling and 23 for uh bearing the

audio

and uh you know that's kind of the high

level now I'm talking domestic

um ISD and PRI here for United States if

you're in a country that uses E1

Services you're going to do ISDN PRI E1

and you know pretty much the same thing

you've got additional time slots

um but if you're living in the E1 world

you understand those things already so

no need to readdress it here so let's

jump in a little bit here and talk about

the scenario we're going to mock up like

I said before I've got a Cisco phone in

fact let me draw a switch out here you

know here's a switch and you know I've

got a Cisco 89 45 phone you know and

he's extension you know 2001.

and he's going to make a call

to the pstn

I've got my PBX configured with some

dial plan logic so that we'll use eight

as an outside access code and then we're

going to dial you know a seven digit

number

so whatever route plan entry you want to

use to do that

and I'm not going to go through that

because it's a common issue thing we'll

talk about that in cipt1

but we're going to send the call to the

voice Gateway we're going to match an

inbound dial up here we're going to

select an outbound dial up here that

we're actually going to build on this

video it's going to be dial up here 500.

dial Dash Pier 500 pots

and we'll walk through that

configuration and it's going to send the

call out this ISDN PRI circuit and I

happen to this evening have a ISDN or

you know a PRI test set connected to my

voice Gateway but if you're using an add

train Atlas 550 or a Cisco router is a

psdn simulator maybe you're lucky enough

to have a real isdnpri at your disposal

you know those things will work

identical to this so you know pick your

tool of choice you know I'm just using

the test set tonight because it was with

an Arms Reach of me and uh

so there you go so here's what we're

going to build we're going to walk

through the Soup To Nuts I'm going to

show you some debugs I'm going to show

you how to validate isdns working and

should leave you prepared to support

ISDN in your environment and be ready

for the test

so let me pull down a putty session here

and we're going to take a look at this

router I've got this is my real pstn

Gateway this is what I'm using in my lab

for access to the psdn and tonight it's

going to be a simulated psdn across this

ISDN PRI circuit if we do a show

controller t120 I want to show you we've

got a T1 and physical layers and service

you know we're not taking any errors the

circuit's clean

um but you know it's just a plain old

data T1 at this point in fact if I do a

show run you know you're not going to

see anything you know PRI related you

can look at the T1 it's configured with

the zsesf for its Framing and line

coding standard T1 stuff there but let's

go ahead and make this T1 and ISD and

PRI now the first thing we have to do on

our voice Gateway is Define globally in

ISDN switch type and I'm going to say

ISDN switch type question mark and I

want to show you a few options uh

primary you know four and five ESS DMS

100 primary ni for National ISDN these

are choices that are common here in the

United States if you're you know in the

UK you may be using the net five Etc so

lots of options here we're going to go

ahead and use primary Dash ni for hours

now depending what version of iOS and

model of Hardware you're using you may

also be able to set these

um yeah there's been some changes over

time some places you could send them

both globally and in the uh the serial

interfaces you know some didn't quite

have the same level of support I'm going

to set it globally because I only have

one circuit connected here if for some

reason you have multiple pris from

multiple providers and I'm not going to

get into all the other things you're

going to think about relative to T1

timing and all those sorts of things but

if you have that you may have different

switch types to play with so set one

Global and then for the Oddball set it

on the port but I'm going to set it

Global and it's going to meet the leads

for this exercise so we've set the ISDN

switch type now we're going to go to

controller

t120 and I'm going to say PRI group time

slots

one through six I don't have enough dsps

to really bring a whole PRI up maybe I

do but yeah I don't need to so I'm just

going to pick six of them on this device

and hit enter and that right there has

defined an ISDN PRI on that controller

now check this out if I do a show run

I'm going to navigate down to a bit of

code here that has just been created

you'll see this interface serial two

zero colon 23

. again we've got ISC and switch type

primary ni ISDN incoming voice voice

that is a virtual interface that was

created for our d-channel or our

signaling channel for ISDN on this T1

and uh you know that's you know kind of

what we're going to point at with our

dial pair so we've got ISDN it's

actually in service now so if I want to

do a show ISDN status this is a great

command just to give you a high level

view of how ISDN is doing on your router

so we're going to see ISDN serial 2 0

colon 23 we're going to see the

interface ISD and switch type is primary

ni layer one is active because the T1 is

up and in service you'll see the the

state of layer 2 is multiple frame

established if you see multiple frame

established on an ISDN PRI you are

golden that means that the D channel has

negotiated

um with your hardware and the circuit is

up and in service and ready to place and

receive calls layer 3 status we've got

zero active calls obviously I'm not on

the phone right now because I'm talking

to you guys but uh you know that's

pretty much the the what's going on now

if for some reason this circuit we're

down in fact I'll go ahead and simulate

that by unplugging my test set right now

if I run a show ISDN status you're going

to see in this case the T1 is down so

layer 1 status is deactivated and layer

2 status is Tei assigned so you're going

to know something's up I'll go ahead and

plug it back in here and awaiting

establishment and run it again multiple

frame established so PRI is good and

ready to go now we don't have any dial

plan right now in the voice gateway to

support this PRI so here's what I'm

going to do we're going to go config T

and I'm going to say dial peer voice 500

pots

and I'm going to say destination pattern

8 dot dot dot dot dot dot dot so seven

digits and that's not an overlapping

pattern in my dial plan so I can get

away with a a simple dial up here like

that I want to say port 2-0

and I think I need to go colon 23. there

we go so we've pointed it at our D

Channel

and really that's enough for an outbound

call to take place so check this out

ISDN has two useful debugs show ISDN

whoops I'm sorry debug ISDN question

mark q921 and q931 if I do a q921 debug

and go term mod you're going to see

layer 2 messages for the ISDN circuit as

they occur and uh you know a lot of this

is going to happen you know when I bring

the thing into service but you know

we're going to see some you know

transmit received data here Etc

um you know good indication of what's

happening at Layer Two I'm going to go

ahead and turn that off and show you a

debug that is much more frequently used

for me anyway and then this debug ISDN

q931 now q931 this is my layer 3D bug

this is going to show me what's

happening it's going to show me the call

setup the messages the who's calling

where am I calling all the good stuff so

let's sit back and let's make a call

from my call manager it's going to hit

the Gateway it's going to match that

outbound dial up here obviously it's

going to match an inbound dial up here

but you know that's beyond this video

we're going to match it out on dire pair

that we just created and it's going to

ring the phone and you should hear it

ringing here on the desk so I'm going to

pick up my phone

and I'm going to say eight five five

five one two three four

and you can hear it ringing you can hear

my test set here ringing in the

background I'm going to go ahead and

answer that call it's going to take me

just a second here and hit answer

and answer speech here we go that call

is now connected in fact if I get a

little closer to my phone you may hear a

slight echo in my voice so calls

connected in fact you can see a

connected State on the screen

um and in fact I'm going to go ahead and

hang it up here in a second because I'm

going to get a heterodyte if I don't

um but you know right there that's basic

ISDN PRI we placed a call so calling

party number two zero zero one that's

the extension on my iPhone numbering

playing and typing uses unknown unknown

so that's correct that's what's showing

up here

um we're going to see the called party

number of five five one two three four

that's the digits that my dial peer

passed to the pstn and you know that's

exactly what I wanted as well

um in fact that's a good example you'll

notice my dial up here said eight dot

dot dot dot dot dot why didn't I pass

the eight I didn't pass the eight

because

I automatically strip the specific

digits when doing a wild card match so

only the dots got passed which is

exactly what I wanted so that is a

perfect example of an ISDN PRI call that

was successful in fact you'll take a

look down at the bottom here you're

going to see the disconnect cause normal

call clearing so you know that is

fantastic call lasted 15 seconds we've

got a lot of good information here we

brought up an ISDN call now let me show

you a few other things about ISDN you

know we've shown you how to take a look

at the ISDN status I want to show you a

show istn service command show ISDN

service this shows you the state of the

individual Bearer channels now remember

I only program six of them because of

you know the limitations of DSP

resources on this Hardware

you'll see that the channel State and

the service state for those six B

channels is reflected you know the

channel state is idle

and the service status in service I live

in service that's great indicator if you

look at the channel State on the ones I

didn't program it says reserved and the

service date is out of service so that's

you know that's perfect it matches up

great now normally on a full PRI where

you're using all the time slots you know

those will all be zeros across the

bottom which is what you want now if I

make a call check this out I'm going to

go ahead and make that call again I'm

going to say eight five five five one

two three four we're gonna let it ring

it's gonna take me a second to answer

here answer speech all right the call is

up I'm going to run that same command

show ISDN service and you know again

everything looks good no problems we're

connected and uh you know it gives you a

good idea of what is happening now take

a look at channel six we've got a two

instead of a zero so let me hang this

call up and we'll get rid of the

heterodyne the two says busy

so it shows you that it's on a call so

that's awesome so we've got show

commands for ISDN we've got debug

commands for ISDN you've seen the both

of them

um

let's see what other things do you need

to understand about ISD and PRI I mean

really that's the high level right there

um you know mgcp if you're doing mgcps a

Gateway protocol

um and we'll cover this more later when

we get into call manager but understand

that your PRI D channel is actually

going to backhaul all the way to call

manager and I'm actually using uh a SIP

trunk to my call manager from this

Gateway so I don't have to worry about

that I know what I wanted to show you

um there's a cool command when you're

testing out of voice Gateway and it's

called csam this is one of those famous

famous yet undocumented commands from

Cisco csim is a call simulator I'm going

to actually tie a csim start

and I'm going to type the number eight

five five five one thousand

and I'm going to hit enter

did you hear that and you see the debug

I've actually told my gateway to make a

call now I can answer it on my on my

test set there it's connected obviously

there's no audio stream coming because

there's no other endpoint but I forced

the gateway to make a call all by itself

and in fact I'll go ahead and I will

hang that call up here

but it's a useful command

um

and really that's all there is to it you

know use it and enjoy it so we've hit

the high level here this is a very

practical example of ISD and PRI on a

Cisco voice Gateway

um

yeah really that's it I'm gonna go ahead

and stop there because there's no reason

to hit anything else so we're assuming

in this video that you have basic T1

fundamentals down you know we talked

about a lot of those Concepts and time

Division multiplexing and all that good

stuff in the CC enable Ace so

um you know I think you've got a good

handle on what Cisco's going to expect

you to be able to handle obviously

there's more advanced topics and there's

there's other things you can do with

ISDN you know there's something called

end fast or non-facility Associated

signaling and what nfas is is it's a

method that's going to allow you to

share 1D Channel among multiple ISDN PRI

circuits and you know just kind of

squeak a little additional capacity out

there so really that's what you need to

know and we're going to stop it right

here I'm going to say thanks for

watching good luck with your studying

and I will see you in the next video

[Music]

thank you

all right we've made it to the analog

part of the uh the lecture here and we

live in an analog world and there's a

lot of what we do with telephony that is

because of the way things have always

been done and as we talk about analog

Services we're going to talk about some

of these fundamental techniques that

we've been using kind of since the

beginning of time and they're going to

be a major part of the foundation you're

going to need to understand how you know

these environments work so I'm kind of

rambling here this is going to be long

enough that I know even right now before

I've even recorded the other parts this

is going to be a multi-part video I'm

thinking two or three parts so we'll see

how it lands when I get through it but

this is part one and configuring basic

analog voice Services now when we talk

about analog Services there's really

three categories of analog Services

three types of analog Services you're

going to run into fxo fxs and E M now

let me draw a couple of pictures here

fxo and fxs can be described you know if

I draw draw a voice Gateway

and I take one analog line a pots line

to the pstn this interface right here is

called an fxo interface foreign exchange

office

on an fxl interface

voltage is provided to the equipment by

the line so the pstn or the central

office is providing me voltage to the

voice gateways interface port fxs or

foreign exchange station if I were to

take an analog phone and connect it to

this voice Gateway this port is an fxs

Porter a foreign exchange station Port

The Voice Gateway is providing voltage

to my station equipment or my analog

phone now e m this isn't so much you

know a phone to pstn kind of thing as

much as it is but when you start dealing

with pbxs you know so I could have a PBX

you know and I could have the PBX either

connected to the pstn you know or one

PBX connected to another PBX you know

these are the types of places where I

would likely run into e m signaling and

E M either ear and mouth Earth and

Magneto you may have heard some

variations on the phrase there but fxo

and fxs is pretty much where we're going

to spend our time here but you do need

to understand the existence of e m and a

little bit about it so we'll dive into

this as we go

and I'm going to step back

here in a moment and we're going to take

a CCNA voice look at Analog signaling

and analog circuits and how it all works

you know and if you took my CCNA voice

course you'll recognize some of these

slides and then we're going to Deep dive

a little bit so when we talk about

analog signaling we've got three types

of signaling we've got supervisory

signaling which is responsible for

telling us hey something has changed

like hey guess what the line went off

hook you know somebody wants to place a

call you know that or hey we hung up you

know that's a disconnect you know so

those are supervisory signaling

addressing this is about passing digits

um you know where do I want to call to

and then informational signaling this is

you know I'm notifying the user of some

status so

um I place a call I get a busy signal

that you know I get a reorder tone or

you know something like that that's an

informational signaling message

so three types of signaling and when we

deal with supervisory signaling you're

going to see two types mentioned you're

going to see Loop start and you're going

to see ground star and with this let's

take a step back to the CCNA voice

content that we went through and talk

about in the beginning there was analog

I love saying it that way so analog

waveforms or the stuff coming out of my

mouth when I talk you know it's Rippling

through the air and being captured by my

microphone as an electrical signal I'm

then going to send that electrical

signal over the psdn you know pair wires

and it's going to get where it's going

now fundamental concept here for analog

circuits is the concept of tip and ring

tip and ring dates back to you know

manual switchboards when the pstn was in

its infancy and we have used you know a

lot of stuff hasn't changed really it's

kind of sad to say but you know we've

used this technology for some time now

and the way tipping ring works is we use

these two lines to carry our audio

conversation as well as to deal with

some signaling so normally

a tip of the plug is going to be at

Ground potential and the ring is going

to be at a negative 48 volt DC nominal

potential

when you're in an on hook State and you

know you can hit the Wikipedia link

there at the bottom for some additional

information on tip and ring if you'd

like when we talk about Loop start

signaling

Loop start signaling happens when a

handset is lifted and the circuit is

closed so closing a circuit means

completing a circuit so normally when

I'm when you when I'm just sitting there

with a phone and I haven't made a call I

haven't picked up the handset the

circuit is open and it is not closed

which means it is not connected you know

basic electrical Theory there when I

lift the handset

the circuit is closed making a complete

circuit or completing the circuit

and what happens in fact we'll dive into

that here in a second I'm not going to

get into what happens yet Loop start

signaling is the most commonly used

signaling method for dealing with

standard analog phone lines so when I

say standard analog phone lines I'm

referring to things like residential

service or hey I got a really small

business can I have three Bots lines you

know so this is Loop start connections

you know one of the problems in fact the

most significant problem with Loop start

signaling is that you can run into a

scenario called glare and if you think

back to glare you'll remember that glare

is an event that occurs when an inbound

call and an outbound call

try to seize the analog line at the same

point in time so when the co tries to

send me a call at the same time I'm

trying to place a call you know I pick

up the handset and I'm not getting dial

to them but I hear someone going hello

hello and we're connected but you know

my phone never rang and you know I

didn't know they called me and they're

not sure what's going on either that's

glare

um because of the low volume of calls on

an analog or a residential analog

circuit glare is not a huge problem it

happens once in a while big deal do we

really need to do anything about it yeah

not really so you know low volume calls

you know Loop start signaling certainly

can work

business phone systems where call volume

is higher are much more susceptible to

the effects and the negative effects

anyway of glare so what do we do about

it in fact let's let's not say what we

do about it yet let's get into a little

more detail so let's talk Loop start and

I have to give credit to uh you know

some others out there I've seen this

drawing done a few different ways in the

past in fact you know as it was

explained to me in the beginning

how Loop start signaling really works so

let's draw this as we represent a phone

and we'll call this the calling party

I like that phrase and then we will draw

a CO switch

and really this isn't necessarily

representative of just a switch but the

psdn as a whole and then the called

party

so part T another phone

and if you remember we talked about tip

and ring

tip and ring

so we've got two phones we've got a line

from each phone to the central office

how does loop start signaling work so

let's start with both phones on a hook

with both phones on hook the circuit is

open you'll notice the tip and ring are

you know they're just kind of floating

out there they're not connected

when you pick up a handset

the phone is going to close the circuit

or complete the connection you know and

it's completing the connection with your

audio you know your microphone you know

there's other electronics here to play

but your audio insertion you know you're

becoming the thing that makes that

connection so this is where we're gonna

obviously send our audio down the psdn

but we're going to close the circuit now

when we close the circuit you'll

remember that with f x o connections

which is what this is

the pstn

is providing Loop current now what's

going to happen is the co switch you

know as we draw let's draw current

c-u-r-r-e-n-t as we draw electrical

current the co switch is going to

recognize hey is drawing current that

means he went off hook and he's going to

assume that I've gone off hook and I

want to dial a number so he's going to

play dial tone to me

so that's how dial tone works now I'm

going to respond to that dial tone by

punching in digits and you know gtmf

tones and those tones are going to be

sent to the COS West so DTMF

and he's going to parse him figure out

where I'm going route my call and

ultimately at the other end

he is then going to sing send a ringing

voltage to the destination phone so he's

going to send ringing voltage

on the tip so we'll just say ring

to this phone

's going to ring

when it goes off hook

we're going to close the circuit again

and our call is connected

so that's pretty much you know the call

setup process with Loop start signaling

now what you'll see here and and this is

where the glare issue comes from is

there's no way for me as the calling

party to tell the co switch that I'm

ready to make a call is it okay if I do

so

you know or in other words are you

getting ready to send a call to me or is

it okay if I use this line so that's

that's the what's missing that's the

glare component of things so Loop start

signaling it's not going away has some

limitations glare is one of them uh

disconnect supervision you know

knowledge of you know when the phone has

hung up can be another issue you know if

the far end the caller party hangs up

there's no way for the co switch to

Signal my phone that uh you know I've

had you know the other guy hung up so

you know so that I know to hang up but

uh you know I know because he said

goodbye

but anyway

um that's Loop start signaling

Ground start signaling which came about

with pay phones Etc is a way that we

deal with the problem of glare so when

we utilize Ground start

we have a little different kind of

signaling occurring here we actually do

grounding of tip grounding of the the

ring Etc to you know inform you know the

other guy what's happening and we'll do

a sketch here in a minute

um

Ground start circuits are not

susceptible to the same glare problem

that Loop start circuits are because we

have this way of signaling the co switch

I'm ready to make a call and it has a

way of signaling me I'm going to send

you a call so we deal with that problem

so Ground start signal and let's see if

we can do this Justice with a photo here

so groundstar is typically going to be

between a PBX and the central office

it's not going to be you know just a

standalone phone so let's draw draw a

little bigger here we'll say PBX and

then we'll say Co

so again we've got this tip and ring

going on here and we'll say tip and ring

so when

I want to make a call from my PBX

my PBX is going to Signal the pstn that

I want to make a call and I'm going to

do that by grounding

my ring line

so what's going to happen is I'm going

to ground the ring

like that connected to ground potential

zero volts

the CEO is going to detect that I have

grounded my ringing line and he is going

to respond

by grounding his tip boom ground

so we have a handshaking

that occurs and actually drew that

backwards you see the monitoring of it

so I observe

that ring was grounded and I observe

the tip was grounded

so we've got a signaling here that a

call setup is you know wanting to take

place and obviously when you know when

the call actually takes place you know

then we we close this loop again or

close the circuit you know just like we

would you know electrically it's it's

the same thing happening but we've got a

little bit of a handshaking process here

with ground start signaling so

groundstar solves the glare problem

Ground start gives us a way of doing

disconnect supervision so I know when

the other guy hung up because you know

my status is going to go back to the way

it was before I had a call

and that's really it for part one of the

video here we're going to in the next

section we'll talk about DTMF a little

bit we'll talk about call progress tones

we'll touch on E M briefly

um you know enem we've got uh immediate

start Wing start Etc so we'll hit on it

briefly and in the third part we're

going to actually do a little bit of

configuration on fxo and fxs interfaces

and give you an idea what that's like

and what's necessary to make the stuff

actually work so with that we're going

to stop in on part one here of the

analog stuff What's called the analog

stuff video and I'll see you in the next

video and thanks for watching uh good

luck with your studying and I'll see you

soon

[Music]

thank you

[Music]

all right welcome to part two of

configuring basic analog voice services

and uh in the first section or the first

part of this module we talked about

fundamental concepts relative to fxo and

fxs and e m and you know we really dived

in you know pretty deep into the fxofxs

but we didn't touch on the E M stuff as

much as we need to so we're going to

talk about e m in this video and get

through the theory part of it really

there's not a whole lot to uh to worry

about but uh you know I do want you to

have a good understanding of this for

when the time comes that you're taking

your exam and then finally in part three

we're going to get into some

configuration demonstration and and lab

it up and show you exactly uh you know

what it takes to configure these voice

ports so on to part two of analog

we talked about supervisory signaling in

the first section and now as I mentioned

before there's two other types of

signaling on an analog circuit we've got

address signaling and informational

signaling so let's talk about address

signaling the first type of address

signaling that we had was what we called

rotary or pulse styling and basically

which it has your headphone with a wheel

in fact you know you've seen these

before and you know you'd spin the wheel

and you know we could click click click

click click and you know it would count

the pulses and the switch would wrap

your car well that's long gone

thankfully you know that is that's done

and over with and we're living in a DTMF

or a tone-based dialing world now DTMF

dual tone multiple frequency is a

audio signaling mechanism that uses just

like it sounds dual tones and I'm going

to show you a chart here of

all of the buttons on a DTMF keypad one

two three four five six seven eight nine

asterisk zero pound or the hash symbol

if you're in the UK

but anyway what happens when you press

o1

is we play the 697 Hertz tone

and the 1209 Hertz tone

simultaneously and that's why the tones

sound kind of weird you know when you

play them it's not just one noise it's

like there's there's a couple of noises

happening because there are in fact

dcmf5 for another example is 770 Hertz

and 13 36 Hertz playing at the same time

so this is a way for us to send a

representation of a a key you know in

this case we've got

um you know 12 of them in fact there's

some other keys that we don't commonly

use and I'm not going to get into this

crazy you know there's like an A and A B

Etc

but uh you know DTMF so that's address

signaling when we talk about

informational signaling we're going to

talk about CP tones or call progress

tones Now call progress tones

are used to provide some feedback to you

the caller

on how a call is progressing you know

sounds straightforward enough and again

we're going to show you a chart and this

is different from country to Country so

I want you to consider that this chart

is the North American CB tones you know

consult your local local chart based on

the country you're in so when I pick up

the phone and I hear a dial tone

I'm actually hearing a DTMF

you know combination of audio I'm

getting a 350 Hertz and a 440 Hertz tone

now it's continuous so you'll see off

time and on time you know is is listed

as continuous because we just get the

you know and in fact let me see if I can

get one near a microphone here

there we go

so that's the uh the DTMF for a dial

tone now if I place a call

and the call to party is busy I'm going

to get a busy signal let's see 480 Hertz

and 620 hertz frequency now it's going

to be an alternating series of on off on

off on off so the the time between tones

is going to be a half a second

and the time that the tone plays is

going to be half a second

so that's a busy now if you go down to

the reorder tone you'll see 480 and 620

again same tones

but the the tone time is 0.3 seconds and

the time between tones is 0.2 seconds so

that's the fast busy signal we're used

to so you've got the busy and the

reorder tone which we call the fast busy

you've got a ring back you've got the

receiver off hook you know the receiver

raw cook have you ever walked away from

the phone set it down and you hear it

going

it's a really bad impersonation but you

know exactly what I'm talking about if

you've ever experienced that it's a

really annoying shrill kind of tone so

we've actually got four different

frequencies playing at you you know in

very fast um you know fast repetition so

0.1 seconds of on and off time so those

are CP tones they're called progress

tones now when you're configuring voice

ports on a Cisco voice Gateway let's say

I'm configuring an fxs Port I'm going to

configure it to mimic the CP tones for

the country that I'm in so I'm in North

America I'm in the United States so you

know those are the CP tones that I'll

use but I can actually change the CP

tone country

and use somebody else's CP tone so if I

want my dial tone to sound like a a

United Kingdom dial tone then I can do

that by setting the CPU tone and we'll

get into that as we get into the

demonstrations in lab work of

configuring the fxs ports on The Cisco

Gateway so but this is called progress

tones I want you to understand what

they're for and understand that they're

using DTMF

all right so e m

you're not gonna run into a tremendous

amount of e m signaling in the wild

that doesn't mean it's not there

it just isn't as popular these days as

it was historically with you know

traditional pbx's so e m signaling ear

and mouth Earth and Magneto Etc

depending on you know which uh which

version of the the description you

subscribe to is most often used from PBX

to PBX or you know a PBX to an ACD or a

PBX to an ivr you know something where

it's system to system

there are different types of e m

signaling

the most common type of e m signaling in

North America is the NM type 1

while the most common type outside of

North America is type 5. and different e

m physical interface types

utilize either two or four wires for the

signaling Behavior so on an e m type 1

I've got

two wires for audio

and two wires for signaling so I've got

the audio path which is a pair and then

I've got an e

wire and I've got an M Wire now on E M

type 2 I've got an e-lead an M lead

and then I've got an SG lead which is

signal ground and an SB lead which is

signal battery and you'll see you know

different variations on the theme Here

so when I say two wire it's two wires

for the signaling so it's an e m type

one is actually a four physical wires

you know two for the audio path two for

the signal line so you've got four wire

and six wire connections here but we

refer to it as a two wire or four wire

so how's that for confusing you

um again there's some other signaling

types out there it's important to make

note the DNM Type 4 is not supported by

Cisco gateways and you know they're

they're very um very vocal about that

I in the wild have very infrequently run

into e m and you know as a physical

interface in modern systems

um we use e m type signaling

on T1 Cavs a bit and uh we're basically

taking the analog representation and

doing it in a digital way and you'll

learn more about that as you get into

cat stuff but

at a high level

understand this chart understand em type

one most commonly used in North America

and I'm type 5 is most common outside of

North America

and you know know the other types

with e m signaling we have three

different what I'll call start types and

the start types have to do with the

immediate I'm sorry the um the address

signaling so immediate start this is

very similar to Loop start

and basically what happens is with

immediate start the calling party is

going to Cease the line you know they're

going to go off hook by bringing the

elite off hook and then I want to wait

for a period of time now that period of

time at least for you know where I'm at

is 150 milliseconds

so they're going to wait 150

milliseconds and then they're going to

do their thing and use the line so

really there's no handshaking taking

place it's more of a okay here I go type

of behavior so that's immediate start

now just like Loop start there was

really no handshaking taking place so

guess what I'm susceptible to yes I am

susceptible to glare

e m wingstart is by far the most common

e m

address signaling method used and

wingstart basically is going to

give you a handshaking capability so the

calling party is going to go off hook on

the eleague

and then it's going to wait for a wink

or a temporary pulse sent from the other

lit from the other end on its Emily

before sending the DTMF digits so it's

it's kind of a are you ready yes go type

of a signaling so Wing start the most

common

now there's one more and that is delay

start delay start is used mostly for tie

trunks

and the calling station is going to go

off hook on the e-lade and then after a

period of time the calling side is going

to look at the status of the called side

and then make a decision for routing

based on whether the called site is on

hook or not so tight trunks for that

that's the foundation you're going to

need to understand for E M for the C

voice exam you know nothing too crazy

here we didn't get into

um you know anything you know kind of at

the electrical signaling level all that

much you know we did talk about you know

SG and SB and E and M lines and those

types of things but uh you know Cisco's

you know they understand they realize

that this is legacy and you know not

prevalent anymore but uh you know

there's e m so this is it for video two

or part two of modulate

um so we've covered fxo and fxs and

signaling types and we've covered e m

and in video three we're going to go

through and do some configuration of

some fxo and fxs ports and we'll show

you how to program CP tones and and how

to do some basic dial peer configuration

to use in fxo or fxs line

so with that I want to say thank you for

watching good luck with your studying

process I really appreciate you tuning

in there and hanging in there with us as

we go through these modules I know this

is a lot of information to kind of take

in at once but you're well on your way

to preparing for the Cisco C voice exam

so I'll see you in the next video in

part three

[Music]

foreign

part three of configuring basic analog

voice Services we've made it and uh you

know I'm just gonna jump straight into

it here and give you guys kind of the

what is what is what is so we've talked

about basic

um fxo and fxs logic and Theory

now it's time to do some programming so

you know my lab is always changing and

I'm always adding a piece of Hardware

removing a piece of Hardware Etc and um

you know I'm not even sure which Gateway

I'm on right now sugar

we are on a 3725 you'll see that I've

got two voice fxo interfaces and two

voice fxs interfaces installed in fact I

left them right here

so we're going to demonstrate

provisioning of both so I've got an

analog phone an fxs interface here

and I'm going to try to kick my gain up

just a little bit so that you can hear

the sound it makes when I'm holding it

up to the microphone it's a little tough

to capture

but uh let's go ahead and configure

a fxs port on a Cisco router now the

first thing you're going to find out is

there's very little to configure for

basic fxs functionality

I'm going to do a show around here I'm

kind of show you what's going on this is

actually my psdn Gateway

that I'm using

and I've actually got

an fxo port in it

connected to uh what do you call it a

MagicJack plus and I'm using that as a

you know it's a SIP trunk basically and

I get pstn access from it for crazy

little money comes in great for lab

purposes

but I've also got a pods phone connected

here to Port one one zero and you'll see

voiceport one zero and it's got

absolutely nothing configured and really

you know I'm if I pick the thing up in

fact let me do a debug here uh term mod

debug

oh which one do I want to use

um vpm all

I'm going to take it off hook

and you're going to see

the uh the results of me taking it off

up I'll go ahead and put it back on hook

and you'll get a little debug output

there so I've got a phone hooked up in

fact if I want to make a call I can in

fact let me kick this gain up here just

a little bit and we'll make a call

and you can see exactly what happens

and actually for whatever reason that

call wasn't routable

but uh you kind of get the point here

that I've got an fxs line connected to

an analog phone and I attempt to make a

call and I match dial piers and you know

everything that's supposed to happen

happens

so

yeah it's just really simple now let's

talk about the unique stuff on an fxs

port

um we talked earlier about CP tones

let's do this let's go to the voiceport

config t

voice

Port what did I say it was one one zero

CP tone question mark

I've got a list of locales and I can

select

the locale

for wherever I want this to seem like it

is let's go ahead and pick Z

[Music]

let's see here I'm looking through the

list here

how about the United Kingdom CP tone GB

now when I take the phone off hook let's

listen to it

here goes

actually that sounds like the US let me

find one that's kind of weird

and I should have done this ahead of

time to see what I like but uh

here we go it's kind of a weird one

a little different off hook tone there

let's see what uh the Pakistan

cp10 sounds like

here we go ready

a little different because it's gotta be

a really really goofy one in here I

wonder what Zimbabwe sounds like whoops

config

voiceport

cp10 ZW let's see what Zimbabwe sounds

like

another hum

once you've heard one home you've kind

of heard them all I guess but anyway I'm

changing the call progress tones by

using the cptone command we'll go ahead

and set it back to us

because that's where I'm at and by

changing the call progress tones we're

going to change the DTMF behavior that

occurs other things I can do is I can

manipulate ring Cadence using the ring

Cadence command and you can see you know

patterns 1 through 12 you know two

seconds on four seconds off or one

second on five second off or one and a

half second on three seconds off so we

can change how this fxs interface rings

in fact

[Music]

um

now I'm going to stop there I don't want

to go too far off track so that's a

couple of cool things you can do on an

fxs Port let me show you a few other

things that you can do on an fxs Port we

can configure

um comfort noise

and we can turn that on or off

no Comfort noise

we can mess with

um

oh what's a good one here I'm just

staring at it let's see what jumps out

at me

yeah there's not anything really jumping

out at me I mean fxs ports aren't all

that exciting the fxo ports are where we

can really do some cool stuff but uh

let's see here here we go I know I know

one thing I should show you signal

this is where I would specify Loop

starter ground star so on this fxs port

if I wanted to behave in a ground start

fashion I could do that or obviously the

default is Loop start let me show you

what we've got going on here show Voice

show voice port

one one zero again this is that fxs line

you can see that I'm fxs one one zero

you can see that we're in the UP State

you can see that our

let's find it here

here we go

or signal type is Loop start

you can see

that our ring Cadence is currently

defined by the CP tone selection

and so a lot of useful information here

um again like I said before it's fxs

there's not a whole lot to get excited

with now fxo

a lot of cool stuff you can do with fxo

lines so I mentioned before that I've

got an fxo line connected to a

MagicJack Plus so basically I get a

um you know what would be a pots line

you know like I ordered a phone line for

the Telco

and I plug it in

to my Cisco router on the FX support

and if I do a show run

we'll show you it's going to be this um

voiceport one one or one zero zero

and voiceport one zero zero

is configured with something we call

plar private line automatic ring down

and what plar does

is

when someone calls me on this line and

it rings

the line will automatically answer on

the Gateway it'll go off hook in

response to that incoming ring

and it will then place the next leg of

the call to a destination of two zero

zero one now that's going to have to

evaluate my dial plans so so

appear no not Shadow plane show

dial pure voice summary good command for

that

you can see all my patterns and I've got

a four digit pattern pointing

to my call manager which then Rings my

phone so if I actually were to in fact I

will do that here in just a second call

I'm not going to say this out loud

because everybody will be calling me

here in the lab 614

I'm going to hit the call button and you

should hear my phone ring

and the reason for that

is because of blar

actually I'm sorry it ran rang through

to voicemail but

no worries there but anyway

um you know I've got inbound calling

outbound calling on this FX support

now

this particular FX support is a loop

start connection but just like the fxs

port yeah

voice

support I always forget whether there's

a dash or not one one nope one zero zero

question mark you know a lot of the same

settings

um we would go to

uh let's see here signal

you know and I could again Ground start

Loop start pick whatever I wanted

and uh

you know pretty straightforward stuff

we've got dial piers

that will

send calls to a given port

in fact it'll take me a second to mock

this up but I should probably do it for

you let's see here

show dial

peer voice summary

let's find a good one how about 300 dial

pure voice 300 parts and we're going to

say destination

pattern

5000 okay and then I'm going to say Port

one

zero I'm sorry one one zero

this

is my FX export

so if I were to

place a call

to Port 5000 or extension 5000 I would

evaluate and match a dial pair and you

can now hear the phone ringing here in

the background

because I made that call we'll go ahead

and answer it and hang it up

and that'll

finish up the call simulation there but

you know that's how I point a call

to a voice Port so if I've got a fax

machine you know and I want to point

adenis

to

you know the port where the fax machine

is plugged in I can do that

if I want to

um

you know have a plar connection maybe I

want a phone to be a hotline phone a hot

button phone I can go to an fxs phone

voice port in fact I'll do it just for

the heck of it voice

Port one one zero connection plar

and we will say what's a good number to

use how about 2001

I will pick the phone up

and I didn't dial anything I just picked

it up and now you can hear in the

background an IP phone ringing so I

basically made that a hotline phone or a

bat phone

so that if you pick it up it

automatically makes a phone call whoops

config T voice port 110.

let's turn floor off and now it's back

to a normal phone again

there's really not anything else I need

to show you about fxo and fxs ports

you know they really are that

straightforward

and

you're going to use a lot of these

you're going to find that you know

hanging fax machines off of

a router

is very convenient

and you're going to find that

um you know in a warehouse you may have

a run for a phone that's in excess of

the 328 feet or 100 meters of ethernet

and you're like yeah let's just run up

you know a twisted pair out there cat 3

for a a phone and put an analog phone

out there run it to a Gateway all of

these things are very much things we do

day in and day out in production

environments so configuring fxs and fxo

at least basic configuration something

you definitely want to be familiar with

and uh you know I showed you a good

debug you know the debug vpm let me show

you a couple options there debug vpm

question mark you know we've got all or

we can look at DSP information or you

know we can narrow it down to a specific

Port you know very very useful commands

so I'm going to stop there you know this

was meant to be a really short and sweet

lab on fxon fxs

and hopefully that is giving you a

little bit of an understanding of how to

configure fxo and fxf

if you've got access to some e m

interfaces you're going to notice that

the configuration is pretty much the

same you know you're going to specify a

type

you know so whether type 1 or type 5 and

you're going to tell it whether it's two

wire or four wire Etc but other than

that I mean configuring these things is

pretty much exactly the same so

with that I think we're pretty much

wrapping up with

um analog Services you know there's a

couple of other analog things out there

you may run into you know there's these

things we call cama trunks which are

used for 9-1-1 but again they're they're

configured exactly the same you know you

may find that you know there's a bit of

anti-mapping that you do you know some

some unique stuff for camera drugs where

I'm from you know I don't run into

camera trunks really for the most part

we send our emergency calls just down a

PRI or a pots line so

it's one of those things that uh you

know kind of varies from region to

region

you may find yourself using what we call

a did trunk or a direct inward dial

trunk now we've already talked about

configuring pris and you understand how

Venus you know is presented you can have

multiple numbers point to a single

physical circuit et cetera

analog did trunks

kind of smell like a PRI and in fact

that you can have multiple numbers

writing or you know using the same

physical facilities

but one thing that's kind of backwards

about an analog did trunk is unlike an

fxo line

where the co switch provides the ring

voltage or you know the line voltage

on an fx

on a did analog line it's actually

backwards it's more of like a PBX

functionality

where your Cisco router

is going to provide the power

so kind of a weird service I've seen it

used for facts in the past but it's one

of those things I've only run into a

time or two in the last decade so we're

not going to dig crazy deep into it

because you'll probably never see one

and if you do you know have fun it's

really not not all that different from

everything else that we're doing so

we're going to talk a little bit about

um T1 Cavs and how Kaz works in the next

video

and you know we've already talked about

PRI so I think we've got that covered

pretty well

um we'll go ahead and get into some

additional debug commands and

troubleshooting commands that may be

useful with you we'll talk a little bit

about cross-connecting DS zeros which is

kind of cool being able to use some of

the multiplex trunk cards and uh beyond

that we're ready to start talking about

digital signal processors so with that

I'm going to leave you to your Labs

please take the time set some of this up

do like I do get into it and just poke

around you know make a phone ring

Implement you know one feature then turn

around and you know rip it out and do

something different you know play with

the CP tones and see what you know

Uzbekistan sounds like when you call

there see what you know New Zealand

sounds like or you know China or you

know any of these other locations and

but have fun with it you know it's

important that you play with the stuff

listening to me lecture is going to only

take you so far reading the book is only

going to take you so far watching my

videos is only going to take you so far

you really need to get into the gear get

your hands dirty and play with it so

with that I'm going to leave you to it

um thanks for tuning in and good luck

with your studying and I will see you in

the next video

foreign

[Music]

so

module 9 understanding and configuring

T1 Caz services this is kind of a little

bonus that I want to make sure we cover

as part of the C voice curriculum now

obviously Cisco wants you to know T1

Cavs but you know in all of the ccnp

voice material I've seen

time and again I really don't think that

justice has been done to describing T1

cast and you know at least not in the

video world I think Cisco press has done

a pretty good job with it in some of

their Publications but I want to walk

you through it I want to show you how it

works and we're actually going to place

some calls using T1 cast and really kind

of bring things full circle so let's

talk about T1 cast Channel Associated

signaling or Cavs is it's a data T1 with

24 dsos for voice

but unlike a PRI where the signaling is

handled in a Channel of its own or what

we call common Channel signaling

Cass is channel Associated signaling

which means robbed bit signaling in fact

you'll hear the phrase Rob signaling or

RBS and what's happening and I'll show

you visually here in another slide but

what's happening is we're actually

stealing

um some some space or some data from

frames that would otherwise be used and

we're putting our signaling information

inside them but we'll show you a visual

it'll make a lot more sense now keep in

mind this is a digital circuit however

it's using analog signaling methods so

you can use Loop start or Ground start

or e m

in fact the example I show you we're

going to be using e m wink now there are

three signaling types that are common

obviously there's others you know you

get into loopstar groundstar et cetera

but when you start talking about the e m

signaling types which is what you're

most often going to see on a T1 cast

you're going to be looking at things

like e m feature group b or feature

Group D or the A and A version of

feature Group D and what I want you to

see and really you don't need to

memorize this chart but understand that

different e m feature groups offer

different capabilities relative to Annie

and Dennis and then we show some of the

differences here within the slide so

when we talk about T1 cast framing or

ESF the extended super frame I wanted to

break it down for you and show you

what's really happening so with an ISDN

PRI we've got 23 voice paths and one

Delta Channel that's used for signaling

so in Cavs we're doing things

differently so you already know that a

T1 s24 time slots and you probably

already know that each time slot is

eight bits of data

what you probably don't know what you

may not know is that with T1 cast or

with robbed bit signaling every sixth

frame

is or every sixth yeah it will call it

that every sixth frame is being

um subject

to borrowing

um

the uh a bit for framing so frame number

six 12 18 and 24 or really you know I

can say data for time slot 6 12 18 and

24. is going to be just a little bit

different so a standard time slot we

should time slot 11 there

there are eight bits for voice in that

time slot now 6 12 18 and 24 are a

little bit different we actually drop

the least significant bit or we steal it

we rob it for signaling only leaving

seven bits for voice transport so RBS

hopefully this visual helps you a little

bit

very common not as prevalent as it used

to be

um you know if you're ordering new

service from the psdn from your carrier

you're not going to be ordering T1 cash

you're going to be likely ordering

either ISDN PRI or you know even more

modern you know some kind of a SIP trunk

but you're going to use T1 cast for

interconnections between pbx's and then

in other pbxs and ACD systems and ivrs

so cath is important to understand

I want to demonstrate to you in fact let

me pull it down here from my pstn

Gateway the configuration of a T1 cast

circuit and it's really quite

straightforward it's actually a little

bit easier than setting up a PRI the

debugging is not as intuitive but uh you

know you can certainly handle this so if

we do a show controller T1 you're going

to see t120 is up and actually

yeah everything's I'm taking some slips

but we're not going to let that really

bother us here I'm gonna go config tea

and say controller T1 2-0

and I'm going to say ds0 group and I can

give it a number we'll call it group

zero time slots 1 through 24.

and I can specify the signaling type

now you can see here we've got that fxo

and fxs loop start and ground start and

you know a bunch of e m variants more

often than not in including my example

you're going to be using e m wink start

so we've configured e m wink start and

you'll see as I do that that the E and M

in fact you'll see receive and transmit

you see how they're doing the E and the

M there you know I know I talked earlier

about ear and mouth and Earth and

Magneto you see there receive and

transmit so thank you Cisco we brought

those type slots into service now

we're going to let's see here clear

counters let's see if that thing stops

slipping pictures there we go

um show controller

t120

yeah we're still slipping but I'm not

gonna spend a whole lot of time worrying

about timing on this it's not going to

cause us any problems here in my little

limited lab scenario uh debug command

you're going to use for Cavs probably

the the most significant debug command

debug voice CC API and out

um obviously other debugs are out there

this is the one that's going to give you

the magic juice so I've actually got a

test set connected to my router it's a

TTC t-burg 2209 and I'm going to use it

to place a call

down this Cav circuit and then we're

going to uh you know you're going to

hear it ring a phone in fact let me turn

the ringer up on a phone here nearby

that is going to ring

all right hopefully that's loud enough

you can hear it but I'm gonna go and I'm

gonna say what's the extension on this

phone 2001. I'm going to say 2001

and I'm going to go off hook

actually I did that wrong order off hook

2001.

there we go

wait for it

it should be ringing here why did that

not ring let me try that again

unhook

off hook I may have forgot to build the

dial pair 2001

nope there it goes I just uh had an

error in my dialing so you can hear my

phone ringing in the background you can

see the debug here in fact if you look

at the called number right here called

number equals 2001.

we've got a ringing phone in fact oops I

missed it I wasn't fast enough to answer

it let me call it back again and I'll

actually answer it and we'll let you see

some debugs as things get in service so

off hook

I'm hanging on hook let me hang it up

I'll 2001

there it goes there's ringing in we will

pick it up hello hello hello we are now

on a phone call and I'm gonna hang it up

now you're gonna see the debugs proceed

you're gonna see the disconnect and the

cause and you know everything you would

expect there we'll go ahead and go on

hook

with the test set

and there you go we've got T1 cast up

and in service and working I mean really

it's not you know that much difficult or

that much more difficult

to configure anyway then ISDN PRI but uh

you know just a little bit different

Twist on things so

there you go that's what you need to

understand about T1 Cavs

um

so I think we've really covered most of

the voice interfaces that you're going

to deal with in a Cisco UC environment

you know or at least you know the ones

are going to deal with most often we've

talked about analog fxo and fxs we've

talked about e m interfaces we've talked

about ISDN PRI and now we've talked

about T1 cast so with that I think I'm

Gonna Leave It uh you know kind of stop

right here there's some other things you

can do you know I can go into

um these uh these controllers and I can

actually create some cross connects if I

want and do drop an insert and uh you

know all kinds of things I can do you

know in the TDM world that you can play

with but actually you know what one more

thing before I end this up let me talk

to you about a couple of other debug

commands that are useful here show voice

Port summary

very useful to see your voiceport and

you know kind of what's going on you can

see this two zero zero

you know that's the T1 cast we've got

it's configured a m wink it's up you

know in fact let me make a call and I'll

run that command again to show you what

it looks like when there's a call in

progress off hook two oh one and

answering so now let me just run that

command again

so you're gonna see

um channel one is seized so causing

progress and we go ahead and hang that

up again and hang it up on the test set

there we go

um

other debugs or other other show

commands that are useful so we did a

show voiceport summary we can do just to

show voiceport we're going to get a

little more detailed information so

there's my fxo poor we'll bang through

these and get down to the uh the Cavs

interface where is it at here

I'm looking for it there it is

slot two subbing at zero Port zero

receive and transmit so e n m type of

voice interface is e m it's dormant Etc

um you know you can get lots of

information about Echo cancellation and

you know Etc I'm gonna dig too deep on

that

um

we can take a look and you've seen me

using this one quite a bit but show

controller

and I've been going detailed you know

for like show controller T1 in fact I'm

going to do that show controller T1 you

know we can see you know physical

interface statistics yes I'm still

slipping don't worry about that it's a

lab I haven't defined a proper clock

Source

um show voice DSP and we'll talk more

about dsps in another video with show

voice DSP also useful to see how your

dsps are allocated and what they're

available to be used for

um show voice call summary

um very very helpful in fact I'm gonna

do that call one more time maybe one

more time hell maybe I'll do it twice or

three more times again we'll just do as

many times as it's useful for us and I

want to show you the show voice call

summary here

um oh 2001

there we go it's ringing

show voice call summary

you know you'll see that we have a call

and M connect

so useful there we can do a show call

active voice

show call active voice

and get information you know set up time

connection IDs codec information

um lots of lots of debug related stuff

you know packet counts and you know

number of

um

I scrolled by it here already

let me find it here where is it ah there

we go transmit receive bytes and packets

lots of different information about the

call and the call eggs

and one more here show call history boys

show call history boys

and get you know again detailed

information about the call that was set

up and a lot of these you know you can

go ahead and show you like brief you

know get some summary information so

play with it a little bit play with some

calves

um it's good to go router to router with

gas so if you want to set that up you

know just remember clock Source internal

on one side clock Source line on the

other and you should be good to go and

that pretty much wraps up our lectures

and our videos and our demonstrations on

configuring voice services so with that

I'm going to say thanks for watching uh

sorry I've been kind of dragging it on

here in this video and getting a little

chatty but I think we've hit the high

points and I think you're you're ready

to go as far as the key competencies

that Cisco wants you to have so we'll

see you in the next video good studying

and thanks for watching

[Music]

thank you

all right we've made it to module 10 and

we're going to kind of wrap up actually

in this video I think this section of

the course where we're talking about you

know fundamental concepts and we're

going to wrap it up by talking about

dsps digital signal processors now

dsps do a lot of different things for us

you know there are actually Digital

Signal processors and many components of

a Cisco unified Communications

environment but the dsps we're going to

focus on within this module are the dsps

did it get installed in your voice

Gateway or in your router so let's talk

about the different roles of a digital

signal processor

one of the key functions or

responsibilities of a DSP is transcoding

and transcoding is direct conversion

from one audio codec to another so let's

say for example that I am on a phone

that is only capable of supporting g.711

and the person I want to call is on a

phone that is only capable of supporting

g.729

for us to be able to communicate with

each other our call is going to have to

utilize the services of a transcoder to

convert one codec to another

role of a of a DSP within a Cisco router

is voice termination and then this

really is coming into play when you have

TDM or I'm even going to say analog and

uh IP call legs so when you've got

interfaces from the pstn

and you're going to have a colleague

going to an IP destination so it's the

packetization of voice so obviously PRI

cards you know require dsps T1 cast

circuits require dsps Etc

on Forensic Services both for audio and

for video require the use of DSP

resources and when we start talking

about video capabilities and dsps this

is something fairly new to the the

pvdm3s and we'll talk about pvdm3 boy

that's a tongue twister here in just a

moment

and obviously media termination points

now this is kind of sort of like a

transcoder except no you know no change

of codec is taking place so it's it's

still this termination point where two

audio streams are terminating but it's

simply for the purpose of having two

audio streams terminate to a device so

they can be bridged together and there's

lots of things you can use media

termination points for and you're

learning more about those as you explore

a Cisco call manager within some of the

other videos in this series

we've got some photos and we've got a a

table here I'm going to show you up in

the upper right hand corner that is

actually a pvdm module in fact I think

it's a pbdm3 that's a DSP module now the

actual DSP is one of the chips on the

module but we call the module as a whole

you know a DSP so it's not technically

true you know it's a collection of dsps

or it's a DSP module but you get the

point now the pvdm2 is a legacy DSP

module and it was supported on The Cisco

2800 series and 3800 series of

integrated Services routers now we call

it Legacy but it's still supported and

can be used on the 2900s and 3900 series

isrs but you require an adapter and

that's actually what I've got in the

photo on the bottom right hand corner

shows you the adapter so you take a pvdm

to put it in the module and you can put

it in a pvdm3 slot so thank you Cisco

for form factor changes

with PB Dam twos you had a couple of

different models that were available you

had you know as small as the pvdm 2-8

which would give you eight DSP resources

to use and as high as the pvdm 2-64

which was I'll let you guess yes 64 DSP

resources and to put it quite simply you

know if I am doing let's use media

conversion or you know audio

determination as the example here so

yeah I got a PRI coming in and I've got

23 time slots I need 23 DSP resources to

be able to terminate those 23 time slots

of audio on the PRI so it's really easy

to read these model numbers we'll get

into the Cisco DSP calculator here in

another another slide or two

the uh capability of the pvm2 was it

would do voice and it would also support

some fax capabilities for Cisco facts

relay I'm not going to talk about that

in depth but I do want you to know the

pvdm 3 has video capabilities and you

can actually take a router and do ad hoc

video bridging or ad hoc video

conferencing multi-party conferencing

using the pvdm3 in a router so

pwdm3s we've got bigger steps between

the model numbers but still you know

useful types of quantities pvdm 3-16 up

through the 256.

now how many DSP resources do I need is

a question that you're going to ask

yourself until the end of time it's

always changing and it really varies

based on the situation you're in and

based on the environment that you are

you are working in Cisco has an

application they call the DSP calculator

in fact I'm going to pull it down here

and show you the DSP calculator so you

have kind of an idea what's going on

within the DSP calculator and again you

know we show you the URL here that you

can use to access it within the DSP

calculator I'm going to select the model

Hardware I'm using so I'm going to go

ahead and say 39.45 ISR and we're going

to tell it what iOS I'm using let's say

that I'm on 15 1 4M and it's going to

ask me what cards are in your router so

we're going to go ahead and say that in

slot one of my router I have a v Wick 3

2 mft T1 E1

and it tells me the maximum number of

voice channels that are supported on

that card now what I need to do is tell

the calculator how many sessions of a

given codec do I want to use in fact you

can look here down at the bottom

and Cisco lists low medium and high

complexity codecs and I'm going to have

some notes in the next slide it's

actually a recycle from my CCNA voice

certification class video

where I tell you a little bit more about

this but you know simply the DSP

calculator is going to ask me how many

streams of a given complexity type do I

want and it's going to ask me you know

am I using DSP sharing you know am I

talking PVD m2s or PVD M3s you know I

can hit next fill out you know exactly

what I want to be able to do

and you know the system will then give

me the results

uh maybe we'll see next it's going to

tell me exactly what I need so it's

going to say DSP modules required it

tells me I need one pvdm 3-64 for the

requirements that I fed into the

calculator so a very useful tool

definitely something you're going to

spend some time with if you're working

with Cisco voice gateways for any period

of time and I don't know why my web

browser won't scroll out there we go so

that's the DSP calculator now codec

complexity we and we skipped over it

real fast there and I want to zoom in on

it just a little bit more

and talk about what is codec complexity

so Cisco classifies codex based on what

they call a complexity and really the

complexity is how much horsepower what

kind of resources does it take for me to

process and do what I do utilizing

streams that are you know using a given

codec so Cisco says the g711

g729 a law Etc are medium complexity

codecs and Cisco says that G7 uh oh what

do I want to use g723 or ilbc are high

complexity codecs and back when I was

first doing voice we didn't have a

category called Low complexity and you

know depending on the revision of the

test you're in and what you see and what

kind of documentation you're reading you

need to keep that in mind so the DSP

calculator in fact I've still got it

open let me pull it down here and show

you a little footnote at the bottom here

the DSP calculator is considering g711

as a low complexity codec it that wasn't

always the case we used to consider g711

a medium complexity codec so keep in

mind that uh that there have been some

changes over time so when you're asked a

question about you know low and medium

complexity codecs you know think about

you know how the question is structured

so you know whether you know if it's

talking about g711 is it medium or low

so we're going to stop there and pause

for a break and I'm going to pull up a

router and we're going to go through the

process of configuring some DSP

resources both on a Cisco router or

voice Gateway as well as within Cisco

unified Communications manager call

manager

dsps are accessed on a router using the

skinny call control protocol so we'll go

through some skinny setup we'll build a

DSP form of the router and show you how

to allocate resources for both

conferencing and transcoding purposes so

we'll be right back and give you a walk

through of how to do those various tasks

and then that'll be it for this video as

we talk about digital signal processors

all right so we're going to go through a

demonstration of some DSP configuration

and I want to first start talk uh start

talking about codec complexity so if you

need to configure a DSP

to

support different kinds of codec

complexity

you're able to do that and you've got

different types of complexities that are

supported and in fact I'm just going to

jump right into this we're going to go

config t

and I'm actually on a 2800 series ISR

here and I've got a pvdm 216 in fact

watch this show diag Pipe again pvdm

let's see if that's useful yeah pvdm

slot zero I have a 16 channel

pvdm2 so pvdm 216. so what I'm going to

do

config t

is I'm gonna say actually I'm gonna show

you the state before I do it I'm going

to say show voice DSP

um

and then I'm going to say detailed

and what I want to show you here is that

currently

we have

you know as I showed you 16 channels so

here are the 16 channels

the DSP is in use for the 5510s

and you know you're seeing all 16 of

them

this is what we call Flex complexity

if I need to be able to support medium

or you know I'm going to start the other

end High complexity codex which lets me

do high complexity or a combination of

you know high and lower complexity I'll

go voice card 0 and I'll say

codec complexity and then question mark

and I've got high medium secure and flex

so I'm going to say hi

now that same command that I just ran

that previously showed me 16

resources available now is only going to

show me six

and you can see that uh you know we're

configured in high complexity mode here

obviously by the the count the number

we've got

if I want to change that

voice card 0 go to complexity medium

and show voice DSP detail whoops get rid

of that

now I'm up to eight

and if you're curious medium complexity

codecs support medium or combination of

things lower than media so medium and

low

um basically it's half the available you

know resources if you were running in

Flex mode

I can go back to flex mode config voice

card zero and codec complexity Flex

and then show voice DSP detail and now

I'm back to 16 channels so it's a matter

of how many DSP channels or resources do

I allocate to a given voice Port now

you're not often going to have to mess

with this

in fact I can't think of a time where

I've really had to mess with this and

you know far back as I can remember

because typically we're running g711

to the Gateway and you know I'm not

doing g729 you know stuff here

you know if I'm doing g729 it's you know

between phones across lands not to a

voice Gateway so you probably won't have

to monkey with that too often but uh you

know certainly the possibility is there

all right now that we've configured

um

uh basic uh code of complexity within

dsps and allocated the proper parameters

there the next thing you're going to do

with the DSP form or with with dsps is

configure a DSP form and the reason

you're going to do this is to either

support conferencing services

transcoding services or both

now

the way you configure conferencing and

transcoding is pretty much the same so

I'm going to walk you through

configuring conferencing and then I'll

let you know what the differences are if

you were going to do transcoding so it

all starts out with a DSP Farm profile

so we're going to go ahead and say

config T we're going to say DSP

Farm profile one and we're going to call

it we're going to actually say whether

it's conference or transcoding so this

is one of the one of the differences

obviously I'm then going to specify the

codecs that I'm going to want to support

so we're going to say codec g711 EULA

codec we'll keep this one really simple

g729 RH

okay so I'm going to support those two

codecs I'm going to define the number of

sessions so we can have maximum sessions

and I'm going to say two I'm going to

support up to two conferences

now I'm going to say

associate application skinny

and I'm going to say no shutdown that's

going to bring this DSP Farm profile

into service so if I want to go out here

I can say show DSP Farm profile one and

you'll see that it's a conferencing

profile I'll get that right here

and you'll see that I've got two

resources configured two are available

and it tells me the maximum number of

conference participants are eight so if

you stop and think for a second

I can have eight participants

in two conferences which equals 16

resources

so no shutdown

now we're going to go ahead and

configure skinny we're going to say

skinny

local

fast ethernet 0 0.30

and I'm going to say skinny CCM and put

the IPM I call manager 10 10 210.10

identifier one which is my call manager

one

and then I'm going to say Priority One

version and I'm going to give you a

question mark here depending what

version of call manager you're using

you'll need to specify that so that

you're running the right version of

skinny so I'm going to say seven

hopefully everybody here is on seven or

newer and then I'm gonna go skinny and

turn skinny on

now I'm going to associate some things

together here I'm going to say skinny

CCM group one

bind

interface

enter face fa0.30

we're going to say associate ccm1

Priority One

associate profile I'm going to give it a

name I'm going to call it one register

conference is what we're going to call

it

so if I do a show CCM

um not CCM show sccp

you're going to see

that were configured

skinny is up

you can see the address is important

numbers

and you can see

that

we are active and in progress and our

cost code is not registered with call

manager so what do I need to do I need

to go to call manager so here's call

manager

and we're going to we're in media

resources conference bridge and I'm

gonna

hit find and you'll see the software

conference bridge that is default there

it's running on the server but I'm going

to go ahead and add a hardware bridge

and because this is a pvdm2 it is an iOS

enhanced conference Bridge

now I'm going to give it a conference

bridge name and that needs to match

what I put

out here in this associate profile one

line so we called it conference there so

conference

I'm going to put in the default device

pool and tell it it's a non-secure

conference bridge and hit save now check

this out

we're going to go back

and hit find and you'll see that after a

moment it registered with the call

manager so if I do is just skinny now

you'll see our status is active

and we are registered with the call

manager so call manager is able to

using skinny

access these conferencing resources on

this endpoint so there you go

and if I wanted to do transcoding it's

the same kind of thing but on call

manager I would be adding a transcoder

instead of a conference bridge and again

different models of dsps and different

ships are going to support different

total numbers of conferences numbers of

participants per conference Etc and

that's just something you're going to

have to do a little bit of planning on

and look up and do the math but that's

pretty much it for configuring

conferencing transcoding services on the

dsps on The Cisco ISR

so I think that wraps up our module here

on dsps

and gives you the the core competencies

that you're going to need to have to

pass the exam and to be comfortable

working with these things in the field

so

a lot of other things you can do you

know you could dive into you know

tweaking and tuning echo cancellers on

voice interfaces which is another DSP

function you could talk about

um you know go you know way deeper into

this whole codec complexity thing if you

wanted to but uh you know just the facts

and just the important stuff here for

this pass through so we'll talk more

about this stuff later as we go but this

is a great stopping point for this video

as we get into the next area here we're

gonna get out of this basic you know

voice Gateway introduction type of topic

so we're going to get into some of the

basic voice over IP configuration tasks

so you're going to learn about the Sip

and you're going to learn about

um mgcp and h323 and learn about uh you

know just all kinds of of practical on

the Gateway type of stuff so until then

or until next time I'll say see ya good

luck good studying and thanks for

watching

[Music]

thank you

in this module we're going to talk about

the packetizing and transportation of

voice we finally made it into the the

next section within your C voice

curriculum you know we talked a lot

about hardware and routers and weigh-ins

and colleagues and a lot of fundamental

concepts but we haven't really gotten

into the how does Voiceover IP actually

work how did we manage taking this audio

and sending it across the network well

we're going to cover just that in this

video and talk about packetizing and

transportation of voice so I want to

talk to you about the five steps of

packetizing voice and getting it on the

network into where it's going we have

sampling

quantization

encoding compression and encapsulation

and I know I buzzed through this kind of

fast we're going to spend a little bit

of time talking about each of them here

so when we talk about sampling we're

talking about Nyquist theorem Harry

Nyquist was a mathematician who came up

or postulated

that you would be able to reproduce a

signal by sampling it twice the

frequency rate of the transmitter so

basically if I take an audio sample at

twice the frequency I'll be able to

recreate that signal at the far end so

as this pertains to the telephone

Network the telephone Channel range is

roughly 300 to 4 000 Hertz so that is a

very good subset of the human

um you know the human speech range

obviously it's not you know CD quality

Fidelity if we're only sampling 4 000

Hertz or I should say if we're only you

know able to reproduce four thousand

Hertz but it does suffice and it's what

we call Total Quality audio so if you

take four thousand Hertz and multiply it

by two we're going to sample at 8 000

Hertz or 8 000 samples per second and

we're using a process called pulse

amplitude modulation now what's

happening with pulse amplitude

modulation is essentially we're taking a

snapshot of a signal level at a point in

time over the entire analog waveform

that we're sampling

once we've sampled the audio we're going

to go into a process called quantization

and really we're going to to take that

analog waveform and turn it into a true

digital signal so we took the sample and

the values of the sample are then scaled

we know we're Quantum quantizing them we

are taking a value and we're taking a

sample and putting a numerical

representation to that particular sound

and then this value that we use is going

to define the amplitude of the signal

once we have quantized the signal we're

going to then encode the signal so we

take these Pam samples these pulse

amplitude modulation samples and we

encode them using PCM audio or pulse

code modulation now PCM and in fact

we'll talk about PCM just a little bit

here but let's talk about the encoding

process

just a little bit of History here one by

the data is eight bits and when we're

talking you know in a digital world here

we're typically using values of 0

through 255 to represent you know these

these individual bits of data

so let's take the waveform that we've

sampled and let's divide you know 255

divided by 2. let's divide that high end

binary value by two and uh you know when

we get basically a range

that we can represent the audio signal

as it is a PCM value of negative 127

through positive 127. so obviously if

you're looking at a you're looking at a

sine wave you've got the portion of the

wave that dips below the zero reference

and the portion of wave that dips or

that you know goes above the zero

reference line so that's where you're

negative 127 through positive 127 value

comes up so you know kind of the short

and sweet

um you know if you want to really think

of it that way is 127 plus zero you know

as a position plus 127. so lots of

different values that we can assign to

these samples

an example of how this is represented

within those eight bits is we use the

first bit as a sine value a zero is

going to indicate a negative value

follows and A1 is going to indicate that

a positive value follows we're then

going to take the segment in an interval

for the remaining seven bits and that's

going to present the 8-Bit data so I've

got two examples here so a zero with

seven ones it's going to be a value of

negative 127 and a value of one with

seven ones is going to be a value of

positive 127.

the process of compression is the next

step in packetizing this this voice

stream

depending what audio codec you're using

you're going to require a different or

your occupy I should say a different

amount of bandwidth on the network for

each conversation that is taking place

utilizing the g711 codec we get 64 KB a

second and that's you know plain old

uncompressed audio that's Nyquist

theorem math so if we take 8 000 Hertz

samples so four thousand Hertz

times 2 equals eight thousand and we

multiply that by eight bits we then get

64 kilobits per second so that is a

single T1 ds0 worth of bandwidth now we

may choose to use a different codec

within our Network you know perhaps we

have congested wide area network links

and we want to use a lower bit rate

codec across those wide area lengths

well g729 or one of its variants are

certainly a popular choice for doing

that because we only use roughly eight

kilobits per second of bandwidth

GSM you know just for a reference point

here uses 13 KB a second and then if you

really want to get small it's g723

um you know R53 is 5.3 kb per second so

lots of different choices available to

you as a telephony engineer as far as

what codecs you want to use to transport

voice across the data Network

finally we get to encapsulation and

let's talk about packetization rates

there are two different

um sample sizes or I should say

durations of audio that we're going to

packetize and typically it's either 20

milliseconds or 30 milliseconds and I've

got two examples here I've got a g711

example and a g729 example

so we're trying to show you the total

bandwidth utilized so at 20 milliseconds

we are using an uncompressed bandwidth

for of about 80 KB a second for g711 and

if we use 30 male second samples you

know it's slightly less if we're doing

g729 you know that 20 millisecond sample

reduces down to 24 KB a second and if

you want to see the formula to actually

work all this out you take you know

bandwidth's what we're solving for so

bandwidth equals The Voice payload plus

the layer 3 overhead so the IP overhead

plus the layer 2 overhead

and we multiply that by the packets per

second times 8 bits per byte so keep in

mind depending what layer 2 Network

Technology you're using perhaps it's

ethernet perhaps it's not your layer 2

overhead number May Vary but you know

basically you know voice payload plus

layer 3 overhead plus layer 2 overhead

times packets per second times eight

bits per byte

let's talk about transmission protocols

that are used for moving these newly

created packets around the data Network

first and foremost the star of the show

is RTP RTP is the real-time transport

protocol

RTP Works in conjunction with a protocol

called rtcp which is the real-time

transport control protocol you may have

even heard me refer to it I know I've

heard others call it the RTP transport

control protocol because essentially

that's what it's doing

we have crtp which is compressed RTP and

we have srtp which is secure RTP and

we'll talk about all of these in Greater

detail in the following slides

RTP was created and standardized

in the publication RFC 3550

and RTP defines a packet format or you

know protocol for audio and video

transport over an IP network

now RTP rides on top of UDP and it has

Dynamic Port assignment so it frequently

uses ports anywhere in the range of 16

384 through 32 767. now because of this

Dynamic Port range RTP can be somewhat

challenging to Traverse firewalls and

and you'll see that firewalls such as

the Cisco ASA Etc have developed special

techniques to analyze the protocol and

you know poke holes accordingly if I can

be so blunt but so that's RTP

Dynamic Port range and uh 16 384 through

32 767. now one of the cool things with

RTP is that both unicast and multicast

network types are supported so if I'm on

a phone call from phone a to phone B the

RTP stream between s is going to be

unicast however if I'm leveraging a

solution to send audio to a building

full of Ip speakers perhaps in a paging

application I can still use RTP to do it

but I'm going to utilize multicast

RTP adds timestamps and secret numbers

sequence numbers I should say so that

out of order packets can be reordered

inside of the digit or buffer and it

makes sure that playback speeds are

managed you know being that RTP travels

over UDP we want to make sure that you

know if we've lost data we just skip

over you know we don't want to you know

have replays and data out of order Etc

so sequence numbers and timestamps help

us deal with these out of order packets

assuming they arrive quickly enough that

they can land in our digit or buffer to

be dealt with

I'm going to show you

um two um two different

um I'll just call them you know I don't

know how to describe it frame formats is

a great word to describe it we've got

the RTV packet on the left hand side and

we've got the RTP header on the other

side so you'll see that the RTP packet

includes the layer 2 header

the IP header so so presumably Layer Two

headers your ethernet header layer three

header which is the IP header the UDP

header the RTP header which we detail on

the right and then finally The Voice

payload

the RTP header itself contains the

various fields for Flags payload type

sequence number time stamp and option so

we've got again on the left the RTP

packet framing format and the RTP header

format on the right

rtcp the partner to RTP also defined in

RFC 3550 provides for out-of-band

control information relative or

pertaining to the flow of an RTP session

and its primary role in Practical use is

to provide feedback related to quality

of service

again rtcp both unicast and multicaster

supported and some of the values that

rtcp provides to us is bytes or packets

sent round trip delay packet loss and

Jitter and rtcp packets are sent at

least every five seconds so rtcp the

little brother to RTP

now compressed RTP or crtp is defined in

rfc's 2508 2509 and 3545 and the whole

point of crtp is to decrease the size of

RTP udb and IP headers so when you're

sending voice across the network It's

Not Unusual for the overhead to exceed

the amount of data that you're trying to

move now on a land we don't really care

because there's bandwidth Despair and

it's just not a problem but when we deal

with slow links it's something that we

want to consider so what can compressed

RTP give us well it can compress and

decrease the size of these headers to

two bytes without Chuck sum or four

bytes with checksum so RTP header

compression is going to be used not so

much on the Lan but on point-to-point

interfaces and you do need to explicitly

configure this on both sides of these

links so whether this is a a PPP link or

an hdlc link

um you know you've got to configure it

at both ends this is recommended for use

in fact it's a best practice to utilize

this on slow links so when you have a

Network that has links slower than 768k

go ahead and turn on compressed RTP for

those links and keep in mind though with

the bandwidth savings comes a increase

in CPU load so make sure that your CPU

budget accounts for the overhead that's

going to be created by compressing these

RTP headers

RTP or secure RTP was defined in RFC

3711 and provides for encryption message

authentication and integrity and replay

protection for RTP packets so srtp is

going to use AES for its encryption type

we're going to use hmac sha1 for message

Authentication

and we're going to use sequencing as a

technique to prevent the replay of

packets

one more thing I want to talk about

since we're talking about moving data

around the network removing voice around

the network is bad and that's voice

activity detection and this is kind of

the opposite of moving moving stuff

around the network this is not moving

stuff around the network voice activity

detection classifies VoIP traffic as one

of three types you have speech you have

silence and you have unknown

a maximum of 35 percent bandwidth

savings can be realized when leveraging

voice activity detection on a network

averaging more than 24 calls so what's

vad really do well if a voice packet

uh is classified as speech or as unknown

it will be sent across the network if

it's classified as silence it will not

be sent so by reducing the sending of

packets that simply contain silence you

can save some bandwidth now that is

enabled by default when you're using a

codec that supports it however we can

disable this within a dial peer and

frequently in fact I would say almost

always I disable that within my dial

peers because I may be transporting

things you know other than just voice

that you know have problems with that or

perhaps my users just are uncomfortable

by hearing the the very crisp silence

that occurs when vat is at play on a

network

voice voice transport across an IEP

Network and we've talked a little bit

about the transmission itself and you

know TCP I'm sorry rtcp RTP compressed

RTP and secure RTP as well as voice

activity detection thanks for watching

I'll see you in the next video and we're

going to continue talking fundamentals

of voice over IP and I thank you for

watching and I hope to see you soon good

studying

[Music]

thank you

[Music]

in this video we're going to talk about

the first of our Gateway protocols and

it's really one of the more popular

Gateway protocols in fact it's probably

in fact I don't even say probably it's

the first protocol I ever saw on a Cisco

voice Gateway you know way back when in

call manager three days and it is h.323

so we're going to go through some h.323

fundamentals within this video and we're

going to walk through basic

configuration steps for an h323 Gateway

when we talk about h323 architecture I

want you to understand that it's an itu

standard and really it governs

everything multimedia over a land so we

can be talking about voice we can be

talking about video or we can be talking

about collaboration it's a mature

protocol and it is vendor neutral in

fact it's one of the more vendor neutral

protocols out there relative to voice it

is a peer-to-peer protocol and when I

say that it's a peer-to-peer protocol I

want you to understand that a voice

Gateway

um first off doesn't rely on call

manager you know there doesn't have to

be a call manager to have an h323

Gateway

as the Gateway maintains its own

self-sufficient dial plan so if you've

got a network

and the call manager

and a half a dozen sites and each site

has a voice Gateway and they're h323

gateways each Gateway will have its own

unique dial plan that is controlled by

the device itself it

um you know one of the advantages of

h323 is it allows for very specific call

routing rules even more so than within

call manager itself you know within an

h323 Gateway we can make routing

decisions based on both the called and

calling numbers so very flexible

if you're doing srst or survivable

remote site telephony h323 is a great

option for you because the dial Piers

you use during the course of normal

business will be the same dial Piers

that you're using during a survivability

scenario

like I said before there's no dependency

on Cisco unified Communications manager

I could certainly have a voice Gateway

running h323 and another voice Gateway

running h323 build a dial plan to allow

calls throughout between them and be

done without a call manager

one of the features that is supported on

h323 but not on some of the other

protocols is ISDN nfas or non-facility

Associated signaling nfas

is a methodology where you have multiple

ISDN PRI circuits and you share a single

D Channel or Delta Channel

for

you know a combination of multiple and

in fact I think it's like up to six

different pris so you can buy yourself a

couple of extra time slots for voice by

not allocating a time slot for signaling

so mgcp which we'll talk about later

does not support end fast

h323 does

h323 again very feature Rich has 2.38

support for facts and the t38 is a fax

over IP implementation that's very

popular

and h33 has the advantage of supporting

enhanced call preservation so for

example let's say that I have a voice

call in progress between a Cisco iPhone

and at h323 Gateway you know out to a

pstn caller

if the call manager were to experience a

failure the call would actually remain

in progress because the h323 device is

not reliant on the call manager to

maintain the call so we've got enhanced

call preservation

about h323 Network elements for a moment

we have terminals we have devices called

multi-point control units or mcus and an

MCU is nothing more than a conference

bridge for audio or video

we have gateways and we have Gatekeepers

so a terminal that's an endpoint that

could be a video endpoint a voice

endpoint it could be a computer you know

running you know a video conferencing

software Etc we talked about mcu's you

know this can be devices that facilitate

for audio and build Neo multiplexing for

conferences

gateways are just what they sound

they're where we connect things like

pstn interfaces Gatekeepers are kind of

unique a gatekeeper allows you to

distribute dial plan elements or

distribute logic and knowledge of a dial

plan

throughout an h323 Network and we'll

talk more about Gatekeepers and some of

the next couple of slides

with h323 Gatekeepers which are optional

devices I want to I want to you know

stress that Gatekeepers are not

mandatory within h323 they can be used

for h323 dial plan management

Gatekeepers perform a number of

functions including address translation

admission control

bandwidth control

Zone management

call authorization

bandwidth management call management and

the endpoints themselves are going to

register with the gatekeeper you know

when you turn them on and boot them up

endpoints are going to request

permission or what we call admission

from the gatekeeper in order to place a

call

so as we talk about the components of

h323 I want to mention that h323 is what

we refer to as an umbrella protocol and

what that really means is that there are

a number of other protocols underneath

h323 that make up the entire

communication flow h.225

is you know one of those protocols in

fact we're talking here on the slide

about h.225 Raz which is registration

authorization and Status so when h323

endpoints are using a gatekeeper they're

going to use the h225 ravs protocol to

communicate with the GateKeeper

Gatekeepers have lots and lots of

different message types and I'm going to

go through some of those message types

pretty quickly here and you know

depending on your experience with

Gatekeepers you may want to study this

for the test I remember when I was going

through the ccnp voice curriculum from

IC voice exam they drilled me pretty

heavy on this stuff so take your time to

understand this we have gatekeeper

requests Rejects and confirms those are

the grx messages we have registration

request reject confirms we have

unregister request reject confirm

we have Administration request reject

confirm

we have bandwidth request reject confirm

you see in a pattern here we have

disengage request reject confirm

location request reject confirm we have

information requests acknowledgments

non-acknowledgments and responses

we have non-standard messages

unknown message responses

request in progress

the resource availability indication and

confirm

service control indication and response

and finally the admission confirms

sequence

h.225 you know as we talk about h225

when endpoint communication begins and

we

resolve the endpoint address and we try

to set up a call the next thing that

happens is that h225 messages are used

to establish Communications with the

remote endpoint

you've got a setup and set up

acknowledge message a call proceeding

message a connect message alerting

information release complete facility

and progress now one thing you may

recognize here is this is very similar

to the q.931 protocol we use for ISDN

and that's because h.225 was based off

of q931 I'm sorry and I skipped one

we've got status and Status inquiry and

notify messages actually I skipped too

h.245 call control takes place once a

call is initiated so h.225 comes first

next is h.245 call control so we're

going to use 245 call signaling to do

things like capabilities negotiations

for video and audio

we're going to use it for opening and

closing of logical channels for video

audio collaboration

we're going to use it for flow control

and Conference control and I found a

resource on the web you know I don't

like Reinventing the wheel here I found

a resource on the web with Wikipedia

that gives just a phenomenal breakdown

of the h323 protocol and I've got that

URL shown here so please take a moment

go to Wikipedia view the h.323 page and

I think you're going to find a lot of

value add in that in that link

we're going to go through some basic

h.323 Gateway configurations within this

video and I want to walk you through

what is involved in setting up a 323

Gateway

all right in this example we're going to

walk you through basic configuration of

an h323 Gateway we've got a Cisco 2811

router here HQ router you've seen me use

this device a number of times before and

I want to show you just how dead stupid

simple h323 is to configure we're going

to go convictee and in fact before I do

that do show IP and brief you'll see

that I've got just a couple of IP

addresses you know router on a stick

going on we've got Loop Max zero in use

and in fact I'm going to use the

loopback interface here for a moment I'm

going to say interface loopback zero and

I'm going to say h323 Gateway

VoIP bind Source address 10

150.1.1 what I've done there is I've

enabled h323 on that interface and this

is the address we're going to use as the

source IP address for all of our outline

h323 traffic so h225 h.245 and rat

signaling so it's all going to originate

from that loopback interface again

depending on your network layout uh your

choice of interface you know may very

well be a physical interface

we've talked about dial Piers before and

it's necessary to configure a Diop here

to you know make a Gateway in h323

Gateway in fact specifically is a vape

Diop here so we're going to say dial

peer voice 100 VoIP and we're going to

say destination pattern you know I'm

just going to make it you know eight dot

dot dot

and then we're going to say

session Target

ipv4 10 10 2 10 10. so that's my call

manager and guess what that's at h323

Gateway I mean as simple as that you

know I've got to dial up here

um you know in fact just show run pipe

include or pipe begin dial pure voice

you know we'll see that die up here that

I just built up here you have this dial

up here whoops scroll by it here type

your voice 100 VoIP if I had an incoming

call link hit the Gateway and look to

select an outgoing call lag that is an

h323 Gateway configuration in fact by

default a VoIP Diop here is an h323 dial

pair now when we get into sip gateways

you're going to see that it's configured

almost the same except we're going to

specify a session protocol

and that's fine too but uh you know

really turning up an h323 Gateway is one

of the most simple things that you're

going to do now if I wanted to

um

turn off capabilities for h323

you know they're all by default you know

voice service VoIP

but I can

do a shutdown and turn them off now

you're not going to do that if you're

programming a voice Gateway but you

could

um other things you might want to do

which h323 is configure

whether or not the session transport is

TCP or UDP the default for h323 is TCP

but one of the advantages of using UDP

obviously your application the endpoints

have to support it but if I'm using UDP

I'm going to go slightly faster call

setup so if that matters to you you

might consider this but basically what

we'll do we're already Under The Voice

service VoIP menu and I'm going to go

ahead and say

um let's see here h323 whoops gotta get

on the keyboard right

and then let me show you here we go

session

transport TCP or UDP so I could go

UDP

and tell it to use UDP for h323 but I'm

going to go back to TCP which is the

default I've never changed it in

production but you certainly can

if I want a monkey with the h.225 timers

I've got an idle call connection timer

so we're going to be actually I'm in the

right place right now I can say h225

question mark I want to show you all the

things we've got going on here

um the one I'm going to mess with is

Timeout

and I'm going to mess with the

TCP

call idle

and then I can set a value here or I can

say never so if I wanted to change the

timeout

I could do that the default is going to

be 10 seconds

but play with that as necessary you can

do a lot of different tuning with h225

you've got um

you know in a voice class if I want to

configure a voice class let me get out

of here I'll show you that we'll say

voice class

and then I'm going to say h323

and let's give it a tag we'll just call

it Josh can I do that hang on oh it's a

numeric tag okay we'll say one

sorry about that we will say h225

timeout

gcp establish and then we can put a

value you know I can say you know five

seconds if I want

um I can mess with the h2d5 timeout

setup

by going h225 timeout setup question

mark again I've got some timers here

four seconds whatever

and you would then associate The Voice

class with a dial pair or series of dial

Piers if you wanted to use it now we're

getting into some of the you know some

of the granular stuff here but uh what

was that dial up here I had Show run

pipe again dial pure voice

uh we called it

100 okay so config team dial pure voice

100 VoIP and I would say voice class I'm

sorry

I would actually say voice Dash class I

always forget hyphens voice class and

then I'm just going to say

h323

and then that tag I used so one

I wish Cisco was a little more

consistent in tags of numeric versus

alphanumeric but we'll take a numeric

tag there so anyway you know the timers

and the parameters that I tweaked there

would be available to you

now if I want to verify the

configuration of my gateway you know

show Gateway kind of a real easy command

we can show h323 itut version 4.0 ht33

stack version 0.1 the h323 service is up

it shows us that the Gateway is not

registered to any gatekeeper and in fact

I have not seen a gatekeeper in use in a

network and

I don't even know how long they've

fallen out of popularity because quite

frankly there are better ways to manage

dial plan than using h323 gateways and

Gatekeepers

that said they do exist and you may run

into them from time to time now how do I

debug an h323 Gateway I'm going to show

you a command you've seen before it's a

debug you're going to love to hate or

hate to love one of the other debug

voice CC API in out if I've got an h323

Gateway and I'm sending calls to and

from call manager and I want to know

what's going on that bad boys the debug

we're going to use so with that I think

we've covered fundamentals of h.323 and

um I think you're ready to move on to

the next section we're going to talk

about sip we're going to talk about mgcp

and get into some of the other Gateway

protocols

and you know kind of dig into the the

how and the why and the where and how to

debug Etc

and I think that's enough to move you on

to the next video so thanks for watching

guys have a great evening and um you

know good studying I'll talk to you soon

[Music]

in this module we're going to talk about

one of my favorite protocols and that is

sip now I have to admit that in the

beginning when I first began working

with Cisco unified Communications I was

a h323 an mgcp fan and those were the

protocols I used and I would never touch

set

part of that was because of where it was

as a protocol and maturing

part of it was because I didn't know how

it worked

and part of it was because you know

other people told me not to you know

they said 323 and mtcp and you know

these are things you need to be using

well you know 10 years later here we are

and let me tell you my opinion has

changed or I should say my opinion has

evolved just as the Sip protocol has

evolved the widespread adoption or

adoption of it in general has evolved

and uh I think you'll find that it is

absolutely a contender technology for

voice gateways today so let's get into

talking about Sip and understanding sip

architecture so like h323

sip is a protocol

created by a standards body

h323 was an itu standard and sip is an

ietf standard so we've got two competing

kind of sorta governing bodies and their

individual protocols and it was

developed as an alternative to h323 a

lot of the things you're going to see me

say about sip are the same things that I

said about h323 you know they're both

mature and vendoral neutral a sip also a

peer-to-peat pro peer-to-peer protocol

and each Gateway is going to maintain

its own dial plan logic

um there's no extra configuration that

needs to be done if you're using srst

you know the same dial peers you're

using for call routing can be used for

survivability features it's got no

dependency on call manager again you can

do ISDN fast and you can do t38 support

for faxing now keep in mind if you're

doing t38 for faxing that that is

supported with SIP UDP only not with SIP

TCP but that's a little bit beyond the

scope of the C voice exam so don't

stress that too much but if you're doing

fax over IP you'll know exactly what I'm

talking about

let's talk about sip Network elements

there are two types of network elements

I want you to understand and then sip we

call these user agents the Sip UA there

are sip UA clients or what we call a UAC

and those include endpoints such as

phone video conferencing codecs voice

gateways and then there are endpoints or

I should say there are user agents we

call servers or uas which includes sip

registrars and proxies and redirect

servers and location servers

let's walk through a basic sip call Flow

you'll understand once we get through

this how simple sip really is and how a

call Flow works so let's take a uh a

network example here we've got a phone

and a voice Gateway

an ipwan another voice Gateway and a

phone so you'll notice there's no call

manager here and let's walk through the

Sip setup so step one the phone

initiates the call

and the voice Gateway sends a sip invite

message uh to the other voice Gateway

across the IP network

step three we get a 100 trying response

step four we're gonna ring the called

party

step five we're going to send a 180

ringing message from the destination

voice gateway to the originating voice

Gateway we're going to play ring back to

the phone in Step six

when the called party answers

we're then going to send a 200 okay

message back to the originating voice

Gateway we're going to receive an

acknowledgment and the RTP stream will

be set up end to end so that is a sip

call setup now let's say this session

runs for a few minutes and it's ready to

shut down one party hangs up the Gateway

sends a buy the other Gateway sends a

200 okay to acknowledge it and we're

done that is the basic sip call Flow

when we talk about sip addressing sip

addressing is based on internet URL so

the typical format you're going to see

is user at domain and just like other

protocols we have a URL type and our URL

type is sip colon so just like you've

got HTTP colon FTP colon Etc we're going

to use sip colon here are some examples

of sip address types we've got fully

qualified domain names you know sip

colon Josh at how to network.com we've

got e164 address variance sip colon and

the number at how to network.com we've

got some additional information in there

as well and then we've got some mixed

mode examples so sip addressing can

either be learned through DNS or it can

be the responsibility of the proxy

server if you're registered to a sip

proxy to resolve the address for you

now I want you to understand that sip is

going to leverage sdp or the session

Discovery protocol I'm sorry not the

session Discovery protocol the uh hang

on let me tell you exactly what that one

stands for I don't want to lie session

description protocol

um the session description protocol is

utilizing offer and answer messages and

what you'll see there we go got my head

around order there a little bit what

you'll see is we've got a number of

fields here version origin session time

times connection data media and AVP

codecs and all of these Val values are

going to be populated into the sdp

message so inside of the zip packet

you're going to see this sdp message and

this negotiation taking place here I'm

showing you an example of what an STP

message might look like version 0 origin

is Josh we've got some additional

information in there network type

address type address we had a session

equals and I've got a session name

called my session we've got a start and

end time of zero

we've got connection data

including our network type and address

type and our connection address we've

got our media negotiation we're here

we're trying to negotiate a list of

three codecs in fact I'm gonna Circle

them here for you we say first g711

Ula g711 I'm sorry g729 and then GSM so

a couple of different negotiation

options taking place there

let's talk for a second about delayed

offer versus early offer and this is

really about fdp sdp like I just showed

you is used for

capabilities negotiation media

negotiation codec negotiation Etc there

are two modes that you're going to run

into of sdp message exchange there's

delayed offer and there's early offer

I'm going to show you the same network

example here a little bit smaller and

we're going to compare delayed offer to

early offer so in a delayed offer call

step one we're gonna pick up the handset

and initiate the call and send the

invite

we're going to do the 100 trying we're

going to ring the calling party we're

going to play back the 180 ringing give

the ring back tone to the calling phone

and wait for the party to answer

once the answer has occurred here is our

sdp media offer so the person we called

is going to begin negotiating the

capabilities and we tell them here's

what I can do

we acknowledge having negotiated a set

of parameters to use and this is our

media answer we have our RTP stream set

up when the call is done you know one

endpoint says a buy the other one sends

a 200 okay and we're done

so that's that's kind of your standard

delayed offer sdp setup

with early offer sdp is actually going

to happen a lot earlier hence the name

early offer same network we're going to

initiate a call and in the invite

message we're going to send our

capabilities so that is our sdp media

offer we're going to get the same 100

trying the ring the ringing we're going

to get ring back to the calling party

the call will be answered and in the 200

okay we're going to send our media

answer so that is where our sdp

negotiation is taking place and the rest

of the call is the same you've got the

act the RTP stream one guy hangs up the

other one gives us 200 okay now both

delayed offer and early offer support

RFC 3960 ciprally media and early media

basically presents a way for the sender

to start sending audio before the entire

call setup has completed and you can

read more about that on your own it's

not a heavy topic of the exam but I did

want you to understand the fundamental

differences between delayed offer and

early offer

we're going to get into a nipple heron

basic zip configuration I'm going to

show you how to provision sip on a Cisco

voice Gateway and how to do some basic

sip troubleshooting commands that are

useful with some debug output looks like

Etc so with that

um standby one I'll get the router set

up and we'll go through some basic sip

configuration

all right let's go through a basic

example of configuring sip on a Cisco

voice Gateway and I'll walk you through

some of the debugging processes as well

I've created a Diop here on this Gateway

and I want to go ahead and show it to

you show dial

peer voice summary it is dial peer voice

100.

so show Ron pipe again dial pure voice

so dial pure voice 100

I have a session protocol sip V2 and a

session Target ipv4 1010 210 so this

looks an awful lot like an h323 dial up

here aside from the obvious you know the

emitted destination pattern that's there

but what I want to show you is really

the only thing I did to make this a sip

dial Pier instead of an h323 dial pair

is I change session protocol to zip V2

that's a basic in fact that's the most

basic sip configuration on a Cisco voice

Gateway now obviously I need to have a

lot more data in those dial peers and in

the dial plan

to actually have meaningful call routing

but you know Basics are are enough to

get us started let's create another dial

pair in fact what I'm going to show you

a lot of the times you're going to use a

sip registrar for your gateway or your

endpoints so we're going to go config

team and I'm going to say sip UA

registrar

registrar and I'm going to put in you

know 10 10 2 10 10. I always forget that

ipv4 colon V for colon and I've defined

a registrar I can then Define

authentication

username Josh password

networks Rock whatever so I've defined a

registrar I've defined a username and

password these are my digest credentials

and then I can say sip server

10 10 2 10 10. that doesn't want what

ipv4 colon ipv4 colon

no actually hang on what doesn't want

their sip server ah okay I see what it

is sip server space ipv4 colon 10 10 2

10 10.

it still doesn't like me hang on what am

I doing wrong here ipv4 colon

that is exactly

it should be right

take a look here at another example that

I've got

and see what I'm screwing up sip server

ipv4 colon oh I know what I'm doing

wrong

stupid hyphen sip Dash server why

couldn't Cisco error logs tell me that

sip Dash server ipv4 okay you can tell I

don't use this feature very often but

what I want to show you is I've defined

a sip server in fact let me do a show

run let me get down here all the voice

stuff and show you what we've actually

done we've configured sip UA we've got

our authentication digest credentials in

there so our username and our password

we've defined our registrar and we've

defined our sip server now if you look

at the dial peer I created you'll

remember that I pointed session Target

ipv4 and the IP of the server what I can

do in fact we'll go here we'll say dial

pure voice

101 VoIP and we'll say session protocol

sip V2 and I'll say session Target

sip server

so all I did was redirect

sip server to the IP address I'd

previously configured as my sip server

both of those are valid dioper formats

in fact Show run Pi begin dial pure

voice

you know both of those are legal use

whichever one you want but uh that's

really all that's necessary for basic

sip UA configuration now let's walk

through some debugs debug CC sip

messages I'm going to place a call to

this Gateway from a call manager that I

have configured and I'm just going to

dial extension 7500 now it's not hooked

up to a phone it's not going to ring

anything in fact it's not going to work

at all but I want you to see the debug

output so check this out we're gonna go

7 500

and I got all kinds of garbage flying

across the screen we're gonna step back

and we're going to look at it I want to

show you why I love sip so much from a

debug perspective we're going to start

the call

with a sip invite

and I'm calling zip colon 7500 at 1010

210 1 which is my Gateway

so there's the invite I can see exactly

what's Happening Here

there's a 100 trying

now I actually got back a 403 Forbidden

to which I acknowledged

and you know that was the end of it

obviously it didn't exist so you know we

didn't route it anywhere but uh Isn't

that cool how easy is that to read you

can see exactly what's happening

you know I would say equally as easy as

doing a q931 debug

it's it's just dead now it's easy to

read so that's one of the major major

reasons I love Sip and I can deal with

this on a high call volume Gateway

because it's real easy for me to look at

a message like this hundred trying for

example and know exactly what call it is

I can see the from I can see the two

I've got dates and times it's all right

here in front of me so that's pretty

darn cool if you ask me

so anyway that's a basic sip debug

there's all kinds of show commands you

can use and it really depends what

you're trying to solve and what kind of

problem you're trying to look for

but we can do like for example you know

show

sip UA service and validate that our sip

service is up obviously if you turn sip

UA off zip will be disabled we can say

show sipua

statistics

and get all kinds of metric information

of messaging responses and sdp Etc we

can do a uh show situation which one did

I do I did status

did I do status I did service let's do

status

show sippy way status and I can get you

know all kinds of different information

about the user agent configured and uh

you know what's set up here on the sdp

so tons of things you can get out of it

you know show sipua register

register status you know I can get

information if I had a registration to a

uao registration about that

but really that's going to get you

started

um

just darn cool you know I really can't

say much more about it other than it's

just darn and cool so with that you've

seen what you need to know for sip

you're ready to program sip on a voice

Gateway and uh you know you'll learn

more about programming the actual chunks

and call manager when you get into cipt

so I'm not going to go show you that but

it's relatively straightforward but

anyway I think with that we're going to

conclude our discussion to sip in the

next video we're going to talk about the

final protocol I want you to understand

and that is mgcp the media Gateway

control protocol

and uh then we'll talk briefly about

things like facts and modem relay and

t.38

Etc and uh that will kind of cover what

you need to know about voice Gateway

protocols and at that point we'll be

able to move into the next big big topic

area which is Unified Communications

manager Express or CME so stay tuned

there's a lot more fun to come good luck

as you uh study in practice for your C

voice exam and I'll see you in the next

video

[Music]

thank you

in this module we're going to talk about

one final Gateway protocol and that is

mgcp or the media Gateway control

protocol we we started out by talking

about h323 and then moved into Sip and

we talked about how both of those were

peer-to-peer protocols and how the

gateways would you know hold a lot of

their own dial plan and kind of be the

masters of their own destiny from a call

routing perspective which is all fine

and good but uh you know everybody

doesn't want that every situation isn't

right for that so mgcp provides an

alternative where the Gateway itself is

a thin device and the dial plan is

managed by the call manager

as we talk about mgcp architecture mccb

is a ietf RFC origination current RFC

number is 3435 which obsolete the

previous version of mgcp and RFC 2705

and mgcp is a centralized device control

protocol

and unlike h323 and unlike sip mgcp is a

client server-based protocol not a

peer-to-peer protocol

now what does that mean for you within

your router well one of the consequences

is if you're utilizing survivable remote

site telephony or srst you will have

extra configuration to do on The Voice

Gateway

in the uh in the form of dial piers to

be able to support that functionality

so keep that in mind if you're selecting

a protocol to use on a Gateway it's

going to have srsd mgcp may not be the

right choice for that

mgcb has a complete dependency on the

call agent what we call the call agent

from uh you know from the perspective of

this class the call agent is going to be

call manager or call manager Express

not support ISDN non-facility Associated

signaling so again if you're if you're

using multiple pris and trying to share

the D channel for signaling you cannot

do that with an mgcp Gateway

mgcp uses UDP Port 2427 for messaging

from the call agent to the gateways and

UDP Port 2727 for messages from the

gateways to the call agent

there are a couple of advantages in mgcp

that I want to make you aware of when to

keep in mind when you are choosing a

Gateway protocol

one of the big advantages in fact the

biggest thing really that mgcp has going

for it

is that it offers you a simplified

configuration within call manager and

it's centrally controlled that's why I

call it a centralized configuration so

we do this thing called

q.931 back call

if we're using a PRI Gateway

in our environment and we're controlling

it with mgcp and we actually Channel or

tunnel that D Channel we back call that

D Channel all the way back to call

manager so call manager is aware of

what's happening with that mgcb endpoint

mgcp supports q6 supplementary services

so if you're doing qsig trunks between

pbx's mgcp may offer you some advantage

in that Arena

now let's talk about mgcp components we

have endpoints gateways and a call agent

there are two kind of endpoints you're

going to hear me refer to in this course

one is a trunking gateway and that's a

Gateway that's going to connect you to

the pstn

and what is a residential Gateway that's

going to connect you to for example an

fxs port on a router where I've got a

pots phone or

you know similar Arrangement like that

so there's the the psdn facing gateways

those are Trunk gateways or trunking

gateways and there's our residential

gateways which is pointing towards our

analog connections

let's walk through a simple mgcb call

Flow and explain exactly how the

signaling is taking place with this

protocol and I think this will help you

understand what's going on now there are

different control commands used for mgcp

and I'm not going to lay them all out

for you you can certainly look that up

on your own

but I will explain what they do as we

use them within this connection we've

got the same topology or roughly the

same topology I've used for other

examples but I want to show you in the

middle

instead of using a pstn example I'm

going to use a campus example so I've

got my call manager or my mgcp call

agent in the middle

and I've got two voice gateways and

these are configured as residential

gateways so we'll walk you through a

call setup between two phones on an mgcp

residential Gateway

first we send an r q n t message that is

a notification request and we're

basically telling the gateway to watch

for events

on the endpoint so watch the phones

we're going to receive our qnt responses

from the gateway to the call agent

when an event occurs eventually such as

my calling party phone

going off hook

the call agent is going to Signal The

Voice Gateway

with a crcx message or is a create

connection message and again this is

issued by the call agent

we're going to receive a crcx response

and we're going to send a crcx request

again that same create connection

request to the destination Gateway

and ultimately the phone will then begin

ringing

when the phone is answered we're going

to get a create connection response

and we're going to send an mdcx this is

called a modify connection

um command and what we're going to do

here is we're going to tell the gateway

to update the connection parameters

we're going to receive a response to the

mdcx request that we sent and we're

going to nail up the RTP Stream So now

end to end we have an rtb stream our

call has been set up

we're going to use notify messages

to inform

the call agent of events that occur

so in this case we're going to initiate

a hangout

the call agent is going to signal both

gateways with a dlcx or a delete

connection message

which initiates the session termination

and the the gateways will respond with

the dlcx response so a very simplistic

example but gives you a good

representation of what kind of

communication is happening when using

mgcp

we're going to walk through two

configuration examples here

one will be for a residential Gateway

and one will be for a trunking Gateway

so stand by for one moment while I stage

the lab and we'll bring some equipment

up here and walk you through the

configuration of mgcp on a Cisco Gateway

all right residential Gateway

configuration with mgcp I'm actually

going to use a different piece of kit

here because I've I happen to have an

ideal configuration of modules

so we're going to use my what is this a

37 25 38 25 I forget what this is silver

this is a 3725

and this is actually the Gateway that I

use here in the lab as my pstn

connection

so we got some fxo ports on it we've got

some fxs ports on it and some PRI

interfaces it's I think I've got a PRI

in it but anyway let's get to the mgcp

related configuration now keep in mind

like I said before that call manager is

in control when you're doing mgcp it's

call manager call manager call manager

so this is software configuration we're

going to say CCM manager mgcp

we're going to say mgcp to turn on mgcp

we're going to define the call agent

mgcp call agent 10 10 210 dot

uh I'm going to use 80. that's my call

manager 9 box

and I almost forgot service type mgcp

there we go

now let's create a dial pair dial pure

voice 77 pots port

oops I'm getting a little ahead of

myself here application mgcp

uh oh oh I'm on Old iOS

we're going to go service mgcp

because the application mgcp is a

deprecated command thank you Cisco

service mgcp I'm going to say Port what

is my port here one one zero

that's an fxs port

and now we're going to go dial

pure voice

two pots not two

78 pots

service mgcp

and then I'm going to say

Port one one one

now

there are different mgcp package

capabilities you can turn on

and we'll say exit fgcp package

capability

I'm going to say line package mgcp

package capability

we'll say DTMF package

we'll use

GM package and I'm not going to go

through what all of these do you're

certainly capable of looking those up on

your own

a lot of these are defaults

P package and then we'll say mtcp

default package line package so again

most of that's not essential

but uh

we're going to go ahead let's see here

what are we gonna do next

let's go Ctrl Z and show mgcb

you'll see that the mgcp administrative

state is up

the operational state is active

we've got our call agent defined

and uh you know we're not connected or

anything because I haven't programmed

this on call manager

but really

that gives you

know it gives you the

residential Gateway capabilities

so CCM manager this is what you would

look at to see if the mgcp device was

registered with the call manager now my

call manager is not programmed so you

know it's going to say registering

but if I were to add this to my call

manager

in fact I will Let's see we you have to

give it config T IP domain name let's

just go to a lab

mgcp uses

fully qualified domain name so

gotta put that in there make sure that

it matches my call manager

let me log into my call manager real

quick I'm not going to show you this I'm

just going to do it

so that you can see from a Gateway

perspective how this would work

so device Gateway add new this is a 37

25 next running

mgcp next

domain name is going to be the device

name so psdn gateway.cisco.lab

and we'll use my default call manager

group

so on module and Slot one is an nm2v

and we'll hit save

and then we're going to tell it what

kind of cards are in and I've got a

vic2fx oh and a vic2fxs save

apply

and let's see if this thing comes into

service show CCN manager

registering with cm

let me go ahead and bounce mgcp

now mgcp mgcp

show CCM manager

let's see if we can actually get it to

register

I haven't programmed the endpoint let me

go ahead and program the endpoint here

to put a

directory number on it here we'll

program that board as a loop start draw

I can hit save

oh that's right I gotta put it in a

device pool

attendant dn5

.fly okay

registered so now we registered with

call manager so really it's that simple

to set up mgcp as a residential Gateway

now

standby I'm going to clear this config

out

we'll go back and we'll walk through

setting up mgcp as a trunk or trunking

Gateway and we'll configure it to

interface with an ISDN PRI and I'm

actually going to do that I'm on my 2811

so I'm going to hit the pause button and

I'll be right back with you momentarily

and we'll do the trunking Gateway

configuration

all right everybody we're back and we're

on another Gateway here at 29 11 I'm

sorry 28 11. and we're going to

configure trunk side mgcp here we're

going to have this thing controlling at

PRI so config T oops let me click in the

right window here config T we're going

to say CCM manager

mgcp

we're going to go mgcp again and one of

the things I want to show you here

is you can actually monkey with the port

number here

so if you needed to you could do that in

fact the example that I'm following

shows that being done but I'm going to

go ahead and take the defaults

just because

we're going to say mgcp call agent

10 10 210.80

and then we're going to say controller

T1000 which is my prr module

we're going to say PRI group

we're going to say time slots and I'm

only going to configure the first six

time slots here service mgcp oh oh uh

what do I got please configure Network

clock participate Wix here I must have

been monkeying with timing here Network

clock

participate Wick zero

and let's see if it lets me do it now

yes there we go must have been monkeying

with timing earlier

so really right there is all you need to

do

on the Gateway

um

you know we've we've defined

who the call agent is

we've

told the physical interface

that it's going to use mgcp so if you

did a show run here we'll jump down to

where all of that code went

there's your controller T1 and yeah I

was playing with timing I've got a clock

Source done a little goofy there but you

know service mgcp

and then down here

we've turned on mgcp whoops actually I

should show you all of that we've turned

on mgcp we've defined the call agent

so what I can do now

I would pick t i p domain name

cisco.lab so I'm going to go into my

call manager again just like I did a few

minutes ago here

we'll add a new Gateway this will be a

2811

and actually you know what I am going to

show you this one

just because it's

something you may end up doing from time

to time

so let's see 10 10 to 10 80.

let me pull this window down here

bring it down there

and we'll log into call manager proceed

anyway

app admin

Cisco Systems all right here we go

so I'm going to say

device Gateway

add new

I'm going to pick the model Hardware I'm

using this one's a 2811.

I'm going to tell if that it's the mgcp

protocol the for the domain name it's

going to be hq-rtr.cisco.lab

we'll pick our default SAM Group

now

module in slot one

was an nmhtv

two animation no I'm not on an mhtv2 am

I no this is slot zero

um actually you know I don't remember

how that thing's configured so I'm gonna

walk over and look at it stand by one

and I'll be right

all right so H with zero on slot zero so

we're going to go ahead and tell it

we're using the motherboard save here

and it was subunit Zero Vic

if you look two mft T1 that's what I'm

using

so save

and then we'll go into the endpoint here

and tell it it's a PRI

and give it a device pool I'm not

worried about all these details but what

I want to show you is show CCM

and you're going to see again registered

to the call manager let me show you a

couple of the commands show CCM

not show CCM so mgcp

endpoint this is going to list

the mgcp controlled endpoints and you

can see here interface T100

and you can see the six dsos that I

provisioned and placed under mgcp

control

what you need to be aware of is that

mgcp

is

um

controlling not the device

but the ports on the device so that's

kind of the beginning to end what you

need to know about mgcp so we're going

to stop there

and this concludes our Gateway protocols

conversations we're going to do a video

on uh voice call Quality and qos and

just talking about why it's important

and why you need it

then we're going to briefly change gears

and talk about facts I know everybody's

excited about faxing you know it's it's

still feels like it's 1970 right and uh

we'll get into a DTMF relay a little bit

but just some spitting polish types of

topics that you'll need when dealing

with voice gateways

and after that we're going to dive

straight into configuring call manager

Express so you thought you were done in

iOS you're not done in iOS yet

but it won't be long

but after all iOS is the majority of

what we're covering in C voice so thanks

guys I appreciate you hanging in there I

know that this isn't the most exciting

material in the world but it is

important for you both from an exam

perspective and day-to-day

administration of these types of systems

so uh thanks for watching

good luck with your studying and I'll

see you in the next video

[Music]

in this module we're going to talk about

considerations for voice quality and

this is going to be really really brief

I'm surprised that we're I'll be

surprised anyway if we even spend more

than like five minutes on this so we

want to talk about some of the

conditions of concern

dealing with maintaining quality and a

voice over IP network now this is not

going to be a qos lecture we will be

covering ipqos in very granular detail

later in this course and uh actually I

kind of wish that it was still a

separate course back when I did my ccvp

we had a whole class on qos and I think

that I got a much better understanding

of Qs Technologies than you get today by

blending it all into C voice but I'll

try to do my best when that time comes

to give you a good solid foundation to

build on so but again this is just a

real summary conversation we're going to

have

some of the conditions of concern we

have when transporting voice over an IP

network one of them is Fidelity what is

the degree that we can accurately

reproduce the audio from the source to

the destination we care about echo echo

happens all the time most of the time it

is not perceptible to the human ear but

Echo has a couple of components we've

got amplitude and we've got delay and

the greater those numbers are the more

perceptible that could become so Echo is

definitely a consideration delay in

itself you know the amount of time it

takes for packets to cross the Voice

network and arrive at their destination

plays into the quality of experience as

does Jitter which is simply variations

in that delay

packet loss is kind of obvious if you're

missing chunks of the data you know

you're going to have missing speech

components however packet loss isn't

really that big of a problem in voice

networks because most of our networks

are healthy enough if we've opted to run

voice on them that's just not happening

that much side tone hearing yourself and

this is something even you know even as

an instructor creating video content for

you it's important for me to be able to

hear myself speak it's important to hear

my own side tone and that's important on

the telephone Network too if you were

talking into a phone and weren't hearing

your own voice back you'd think

something was wrong it would just be an

uncomfortable situation for you so

having that side tone is definitely a

condition of concern for quality on a

voice Network background noise or what

we like to call Comfort noise as a

consideration and there are some

techniques we can use to eliminate

background noise but they're not

necessarily things you want to do and

we'll get into those more as we go along

and and talk about things

latency or delay there are two types of

delay within a network that you're going

to be concerned about there's fixed

delays which are things like coding

delays packetization and serialization

delays and network propagation delays

you know how long it takes to get in a

piece of gear and back out again these

are fixed fixed things they're always

going to take about the same amount of

time

you are going to have some variable

delays and those are mostly with regard

to queuing and digitor buffers and thing

along those lines if I take a look at an

example of a network here I'm going to

walk you through some things so between

a phone and a router

you're going to have

generally speaking you know these aren't

absolutes generally speaking you're

going to have a fixed

coder

delay

when the router has to place the packet

on actually I'm getting ahead of myself

packetization or creating the packet

inside the router

it's going to typically have a fixed

delay see if I can write this small pack

I'll just say pack it so packetization

we're going to have a queuing delay

which is variable

whoops V for variable

because it really depends on how full

your buffers are and how much queuing

you're doing you're going to have a

serialization delay which is typically

fixed we'll just put an S for

serialization which is how long it takes

to put the bid on The Wire

then we're going to have you know a

variable delay through our Wan or our

wide area network

and you know we're going to have a lot

of the same kinds of things on the other

end but the one I want to talk about

most importantly on the on the router

which is a fixed delay is going to be

your digit or buffer

so you can see that

in a network

there are a number of

delays that are going to exist and it's

okay that they exist you know it's the

nature of physics we're not going to

change how long it takes for events to

happen but we need to understand what

kind of an impact they have on our

perceived voice quality and we're going

to talk about how much delay is too much

between 0 and 150 milliseconds we

generally have an acceptable quality

experience

between a hundred and four I'm sorry 150

and 400 milliseconds it's often

acceptable in fact most times it's

acceptable given a reasonable

performance expectation so as a systems

administrator when you're communicating

to your business users what an

experience is going to be you need to

communicate reasonable expectations you

need to set reasonable expectations with

the business community so that they

understand what is normal and what is

not and granted different people will

have different opinions on what is

acceptable and what is not but you know

it's not your job to change it it's your

job to set the expectation so as long as

expectations are set generally between

150 and 400 is okay above 400

milliseconds of delay you know you're

approaching darn near half a second

there that's generally unacceptable so

think about this in the context of

internet connections and putting you

know teleworkers at home with IP phones

you know if you've got somebody with a

high-speed cable modem connection and

they're using the same internet service

provider at home in the same city as you

are at the office generally your latency

is going to be you know pretty low

because you're on the same network and

it's going to be a pretty amazing

experience it's going to work really

well however if they've got you know

they live rural you know somewhere you

know out in you know an hour outside of

Cleveland Ohio where you know there's

you know a gas station every 30 miles or

something you know and all they've got

is a satellite internet connection then

obviously you know latencies are going

to impact them in a way

that the more urban

um you know lower latency user is going

to have so you have to really manage

expectations you get into this 400

milliseconds plus time

with specific X you know exception this

is generally not acceptable for voice

performance

I mentioned before Echo is always there

you know it's it's created by impedance

mismatches and audio you know leaks and

all these other things and it's always

there

but what makes Echo a problem is when it

becomes highly perceivable if you don't

notice it it's not a problem and like I

said before the things that create

problems with Echo is amplitude and

delay so the more delay there is and the

greater the amplitude the bigger problem

you're going to have Echo cancellers

will be required when one-way path delay

is greater than 25 milliseconds and

you'll find that with exception of the

land maybe you're going to have delays

greater than 25 milliseconds

just like when we're dealing with data

networks and I'm going to use fiber

optics as an example here to compare

when I'm designing fiber optic networks

and I'm you know looking at how many

Hops and how many jumpers and you know

loss in patch cables I'm calculating an

optical budget well the same thing is

true in a voice Network and you're going

to calculate what we call a delay budget

and here's an example of one delay

budget and these numbers are completely

fictitious please don't hold me to them

if my coding delay is 20 milliseconds

and I said that was fixed my

packetization delay is 30 milliseconds

fixed queuing and buffering 10

milliseconds variable serialization five

milliseconds fixed my network delay has

some fixed variable components my D

Jitter buffer has a fixed delay you know

these numbers all add up in this

scenario and this is a completely

reasonable scenario for a network with a

Wan you know I've got

145 milliseconds of fixed delay plus you

know 30 milliseconds of variable delay

so I may be seeing 160 to 190

millisecond

um delays on this connection is that

okay sure it's okay you know it's it's

within that range of reasonable

expectation of 150 to 400

um is it going to work pretty good yeah

that could work pretty well as long as

you set the expectations properly so

consider that you do need a calculated

delay budget this comes into play more

when you're going across wide area

networks because on a land things are

just happening so fast it's not as big

of an issue

and I talked before you know Echo

cancellers and I just kind of want to

you know hit home on it here Echo

cancellers will be required when the

one-way delay is greater than 25

milliseconds so do you see what I'm

showing you here about that

it's always going to be bigger than 25

milliseconds I mean it's just so

infrequent that it's lower than that

that you're going to have Echo

cancellers you know in your network

doing things

when we talk about voice and quality of

service there are a number of different

factors that come into play and a number

of nerd knobs we can tune

and things that impact it and this is

just a skim level so don't you know try

to memorize all this stuff you're going

to consider things like header

compression and how can I optimize

bandwidth

um you're going to think about things

like frame relay traffic shaping and

fr12 you're going to think about pscn

fallback options routing calls around

congested links you're going to think

about prioritization and queuing of the

iprtp packets and IP to ATM classes

service mappings and we'll think of qos

at layer 3 and low latency queuing for

voice packets to put them at the front

of the queue we'll talk about things

like MLP and we'll talk about things

like RSVP the resource reservation

protocol for qos which you know we'll

talk about from a textbook perspective

but hardly anybody actually uses because

everybody's dscp based

but I digress I'm going on a bit of a

tangent there so anyway just keep in

mind these things come into play when

planning for and dealing with quality

um and performance and experience within

a Voiceover IP network and that's really

we're going to stop it we're not really

getting into any you know in-depth

lectures here

um but we wanted to give you a touch of

things to come and like I said before

we're going to go crazy deep into ipqos

particularly later three qos and low

latency queuing and class-based weighted

fair queuing and all of those things

later in the course so hang on it's a

lot of fun and we'll geek out a little

bit with it so with that I want to say

thank you for watching in the next video

we're going to talk about facts Services

I know it's your favorite Topic in the

world it's certainly one of mine and I

got my fingers crossed as I'm saying

that but anyway

um thanks for watching good studying and

I'll see you soon

[Music]

thank you

[Music]

welcome to module 16 and we're going to

be doing a deep dive of a technology

that many Engineers kind of run away

from in fact I would say that most of

the voice Engineers I've worked with in

the past have a bit of an aversion to

this topic and the topic we're talking

about is facts

um it seems like facts over IEP or just

facts in general is kind of shrouded in

mystery and I want to try to help

um

explain some of the concepts of how

facts actually works and make it less of

a mystery there's no reason you should

shy away from configuration of facts or

leveraging facts on an IP network with

you know some of the traditional pstn

Technologies it works quite well

sometimes it can be a bit

um a bit confusing because of so many

things going on but let's try to do a

deep dive and give you a good

understanding of what's Happening behind

the scenes with facts so that you don't

develop that same aversion that I've

seen too many times and I'm guilty of it

myself so what is this 1970 so facts has

been around a long time in fact you know

the origins of facts transport date back

to the late 1800s and it seems like you

know as much as technology has advanced

and we've got email and we've got

document scanners and we've got you know

optical character recognition and all

kinds of other things available to us we

still leverage traditional facts simile

technology where we take a picture of a

document

turn it into a a modulated signal send

it across the PN and print it out so

traditional 2.30 facts so what is this

1970 yeah the answer to that question is

yeah find it you know we're still

leveraging this technology and as much

as you might wish it to go away it's so

widespread that it's not going to so

let's take a look at our Voiceover IP

network shown here on the page and this

will give you an example of how facts

might be integrated into your

environment and obviously there are

variations on the theme here but we've

got an ATA device connected to a switch

that has a fax machine attached we have

an analog Gateway that could be a vg224

or you know some other variation of a

analog voice Gateway and we've got a fax

connected to that obviously we're going

to have fax machines on the pstn that

are you know calling us and originating

inbound connections so these devices are

going to exist on your network let's

talk about the different ways we can

utilize facts and how to make this an

asset not a challenge for you

we have three methodologies we use to

move facts around an IP network we have

what's called facts relay fax

pass-through and fax store and forward

now I'm going to talk about some of my

favorite methodologies as we go through

here but we're really going to Deep dive

into how each of these work and give you

an understanding of the fundamental

concepts

I'm going to start with facts pass

through and this isn't really where most

people start most people start talking

about fax relay but I'm going to start

with facts pass through because it's the

most like an audio stream

basically what you're dealing with is

modulated facts data being passed within

the audio codec as though or voice data

fax pass-through is going to use or

start Anyway by using the original

configured codec which you know

obviously should be g711 in most cases

for calls coming from the pstn

fax passthrough will then make changes

to the codec and what we call up speed

the connection if necessary to enable

g711 if by chance you're not using it

and the communications with facts path

through is taking place entirely between

the transmitting and receiving devices

or the originating and receiving devices

aka the fax machines The Voice gateways

and the network are simply treating this

as though it is audio and you know the

data the modulated facts data tones that

we've all heard are going to be carried

in the RTP packets and you'll hear this

referred to as both facts pass through

and modem pass through and really the

difference I don't mean we'll say the

difference modem pass through

encompasses a methodology to allow us to

use modulated data for modems and fax

machines where facts passed through is

kind of a fact specific but let's keep

going and let's talk about vax

pass-through negotiation and how it

works so I mentioned that facts pass

through is sending data modulated data

in the audio stream

now it's not entirely the same as a

regular voice call in that we need to be

able to manipulate the Stream

to disable vad and disable EC because

these things are going to cause problems

with transmission of the modulated data

so let's walk through this fast pass

through negotiation process

again the call is going to start out as

a VoIP call so a call between a Gateway

and a Gateway and you'll see this is our

our internal Network here we've got a

fax machine

connected to uh presumably an fxs port

on a voice Gateway

an IP network in the middle another

voice Gateway and another fax machine so

we start off with a VoIP call and the

faxes you know because this is a voice

call the faxes are going to initially

begin to exchange their T30 data or

their T30 signaling from end to end

now what's going to happen is we're

going to wait as the originating Gateway

the Gateway there on the left we're

going to wait to hear something called a

said tone

and that's a called terminal

identification tone and what it is is

it's a 21 Hertz 2100 Hertz tone

that is signaling the beginning of a fax

so we're going to wait to hear the tone

and we're going to send an NSE a name

signaling event message from the

originating gateway to the destination

Gateway

the destination Gateway is going to

accept our NSC request and what's

basically going to happen here is we're

going to change the codec so if I

originally started as g729 and I need to

become g711 that's going to happen here

if I'm already g711 and I simply need to

disable vad and disable AC that's going

to happen here so once that happens

you're going to have the voice call

established again end to end or gateway

to Gateway

and that fax is going to just Traverse

in the RTP Stream So the modulated data

now how do we configure facts pass

through it's actually really really easy

let me show you

a

Gateway here in the lab let's see if I'm

still connected here or if it timed out

yet we're still connecting all right

cool if I go config t

I'm going to show you there's two ways

to configure facts remote and pass

through on a Cisco router you can either

do it globally

or you can do it at a dial peer level so

you know dial pure by dial up here I

prefer to do it globally but you know

you welcome to do this however it makes

sense in your environment

let's talk about global first and then

we'll talk about dial pierces so if I

wanted to just globally I'm going to go

voice service VoIP

and then I'm going to type fax protocol

question mark you're going to see we've

got Cisco we've got none we've got

pass-through and we've got t38

so we're going to do pass through

um

and then if I hit a question mark it's

going to ask me for a codec so I can say

g711 you law

now right there we're done

this Gateway globally will utilize facts

pass through using the g711 ulr codec

now

there's I remember I mentioned modem

passthrough before

it's probably better

if you don't do what I showed you

but instead one small variation on the

theme let me go ahead and issue a no in

front of that and it's going to be modem

passthrough

NSC codec g711 Eula it's going to have

the same effect as enabling facts pass

through but it's also going to give you

mode and password capabilities so

globally you know in fact if I do a show

run on this Gateway you're going to see

up here in voice service modem

passthrough NFC codec g711 EULA and

we're done I mean that right there when

a fax call occurs it will

do pass-through Behavior

so I told you you could also do it on a

per diele peer basis so what you could

do if you wanted to create a Diop here

actually you know let's just look at my

existing dial pairs show Ron pipe again

dial peer voice let's see what dial

Pierce I've got in here uh let's see

dial pure voice 100 VoIP I like that one

let's do it on that one that's my normal

uh pattern pointing to my call manager

so conficti dial pure voice 100 VoIP I

can go again either facts pass through

or modem pass through

NSE

codec g711 your law and that I appear

will use modem passthrough so again

either Global or specific to dial pairs

you can configure it either as fax

passthrough or modem pass through and my

recommendation would be to go ahead and

just do modem pass through and that's it

obviously this is something that needs

to be done on both the sending and

receiving gateways and keep in mind a

receiving Gateway might not be an iOS

device this might be a Cisco ATA and you

may have to follow different steps to

enable pass through or and I'm not going

to get into in this video you know what

devices support what facts types I just

want you to think about it as being a

you know a two-sided relationship or a

peer-to-peer relationship there's the

psdn Gateway but then there's also the

Gateway where your internal fax machine

is connected whatever device type that

might be so we've shown you an example

here of pstn Gateway configuration for

fax pass-through I'm going to hit the

pause button we'll come back do a little

bit more lecture talking about the other

fax modes and we'll do some more

demonstrations all right so I think

we've talked about facts pass through in

enough detail that you understand the

concepts at play now let's jump to the

Other Extreme of what we call

t.37 store and forward facts and then

finally last we'll cover the different

types of fax relay capabilities

t-37 or what we call store and forward

facts is pretty cool and I have to admit

that before creating this video I had

never configured it ever I'd never seen

it configured anywhere and I think a lot

of the reason for that is because it's

just plain confusing looking at Cisco's

documentation you know I've seen so many

iterations of how this can be done and

the examples really aren't that good so

I made it a point to spend the day in

the lab setting this up and really

getting a feel for how this works and

after doing that I think it's really

pretty cool so I'm going to walk you

through detailed directions on how to

configure store and forward facts or

t-37 now we've got two types of devices

with t-37 we have a device called an

on-ramp Gateway and an off-ramp Gateway

and they can co-reside on the same

Hardware if you want them to

you could choose to use one or both

types in your network you've got on-ramp

faxing

which takes a fax coming in and converts

it to an email to be sent to an email

recipient so let's say you've got a

Microsoft Exchange Server or some other

email server in-house

and you've got just a couple of fax

numbers

that are all going to come one place

well instead of putting a fax machine in

your building you can configure your

gateways what we call an on-ramp t-37

Gateway and what's going to happen is

when the fax call comes in you know

presumably on your PRI it's going to

present genus for that destination in

the destinations the fax machine well

we're going to have some configuration

logic in the gateway to send that call

to what we call a MM lip dial peer or a

multimedia over ipdial pair now what

happens is Cisco created a tickle script

or a TCL script that runs on the router

that's actually going to take the facts

receive it turn it into a tiff file and

spin up an email and send you an email

with this file attachment so it's a

do-it-yourself fax server for inbound

fax now if you want to configure

off-ramp faxing where it's the opposite

direction where you've got an email

coming from your mail server to the

Gateway you want to turn it into a fax

and send it out man you can do that as

well I'm not going to cover that example

here but you can support both on the

same Gateway

I mentioned before that it's going to

use Tiff file attachments and that t37's

purpose in life is to enable

Transmissions of facts over email and

we're going to do that using the SMTP

protocol and I said before you know it's

good for small quantities of users you

are going to end up creating a dial peer

per account in most scenarios there's a

couple of ways to wildcard things but

we're going to keep it simple for this

example and this is a validated example

I've built this in my lab

so start here I know some of the

documentation in fact like all of the

documentation I've seen out there on t37

is very convoluted and difficult to

understand they're talking about lots of

different editions and versions of the

tickle script and it just make your

brain in knots so start with this

example so in this device this is an

example of an on-ramp Gateway we're

going to configure

a in fact give me one second and we're

going to switch to pin mode here

pointer options pen perfect so what

we've got is we're going to configure

some information relative to the facts

type so we're going to obviously set our

IP domain name on the device

we're going to say fax interface type

fax mail and this Command right here

depending what mode your router is in

may require you to do a reboot so just

be aware of that and then we're going to

say fax receive called a subscriber

dollar sign D dollar sign so what I'm

going to do is I'm actually going to

capture some information from the call

as it's placed and use this if I choose

later in the process I'm going to Define

my mail server so 1010 210 44. this

happens to be a Linux server that I'm

running with the exam on it as a mail

transfer agent and I can customize the

the subject line of the email that's

going to come to me I happen to make

mine say Cisco t37 facts but you can you

know make it that you've got mail or you

know anything you wanted to say there

and I've got a origin prefix some

postmaster information

some mail from so what we've got here is

we're going to have mail from this

device

and then send mail from username and

we're actually going to put the phone

number of the calling party

in to the message so you'll see who it's

from and then we set some return receipt

commands so that's that's all MTA

configuration stuff

we're gonna have to download and

configure this tickle script you can get

this from cisco.com it's buried right

now within the call manager Express

downloads but just search by version

number the 2.0.1.3 and you'll see the

tickle script in the zip file so I'm

going to configure an application

service on fax and I'm just calling my

service on fax that's a name that I can

I can change if I want and then we point

it to the app fax mail on-ramp tickle

script and the one I was using was the

2.0.1.3

two dial piers

one and two

we've got a pot style pair

and we've got a mmoip dial pair the pot

style appears doing the same thing as

any other pasta appear you've ever used

and we're going to say in this example

that what I've got Happening Here is

we're going to say pstn

I've got a fax I don't even know how to

draw a fax machine there we go that's a

fax machine we've got a fax machine out

on the pstn making a call I've got my

voice Gateway

and my network blah so there what I've

done is we're going to be sending adenis

of six one four five five five one two

one two so when that Denis comes in from

the pstn to our voice Gateway we're

going to match the incoming called

number on the Diop here so we've

selected a Diop here that dial pair is

pointing to the on fax service and we've

got an information type of facts defined

and obviously we support direct inward

dial so we can route the next top

without prompting four digits the next

top is going to be that same destination

pattern six one four five five five one

two one two and what's cool here is

we've got this service fax on VC on-ramp

app outbound just type it the way you

see it there

and we've defined the information type

as facts and we've created a session

Target so Josh at how to network.com and

that is going to be the email address

that I'm trying to send the email to and

the rest of it you know you can look

those commands up if you want to know

what they have to do they're tweaks and

tunes and nerd knobs but this is really

cool so what's going to happen is the

call is going to come in

across the psdn hit our Gateway present

that Denis match

the inbound Diop here match the outbound

Diop here the Gateway is then going to

receive the facts turn it into a tiff

file

and attach that and mime encoded via

SMTP

to see if I can draw there you go to my

mail server

and it's going to show up in my inbox

I've done it it works it's actually

really cool so I would recommend that

you consider this as a possibility if

you have one or two numbers that are fax

machines or that are fax numbers inbound

and you'd like to send it to email you

don't need to go buy you know a ten

thousand dollar third party fax over IP

server to get some basic facts to email

now like I said before this is an

example of an on ramp Gateway this is

going to take a fax coming in and turn

it into an email it is not going to work

in Reverse you can make it work in

Reverse go to cisco.com pull up the docs

and read about off-ramp gateways

but uh I wanted to really show you a

good working model because every dock

that I looked at had this kind of sort

of working but not making sense it was a

pain in the butt to debug and I just

couldn't figure out what was going on

but this sample does work and I did this

on a Cisco 3725 so just as a frame of

reference

so that covers store and forward facts

uh t-37 we're going to get into uh fax

relay in the next slides and we'll talk

about both Cisco facts relay and t-38

Factory lamp so standby and we're going

to change gears and jump into the facts

relay conversation

facts relay is the final option that

you're going to see

as a method for moving

facts around your IP network

and fax relay is going to be done in one

of two Styles either Cisco fax relay or

t-38 facts relay

so Cisco facts relay and in fact before

I go that far let's talk about what

facts relay is so when we went through

the example of facts pass through

we talked about how the audio

would continue you know using g711 to

the end point you know to the endpoint

Gateway or the destination Gateway

and it was an audio stream

and then when we talked about

um

the store and forward t37

we were talking about audio

terminating at the Gateway

and becoming data in the form of an

email

well facts relay is kinda neither of

those

but it's the missing length the thing

that you probably expected in the

beginning and that's data transport of

facts over the IP network to the

destination Gateway so Cisco facts relay

is proprietary

proprietary I'm sorry and t-38 facts

relay is standards based and it's not

Tut based standard and they both do the

same thing

t-38 obviously is what we prefer

nowadays but there are certain pieces of

equipment that you may run into that

only support one or the other so you'll

need to be fluent on how both are used

looking at the same network as before we

talk about Cisco fastc's relay and

negotiation

when the initial call was made the T30

signaling hits the voice Gateway again

we're going to cut through an RTP path

to the destination voice Gateway and

we've got the T30 happening and N so

just like before

we wait for that said tone to occur and

you remember that was the called

terminal identification tone that 2100

Hertz tone

and what's going to happen is when the

said tone is heard The Voice Gateway is

going to listen for the this message

sent by the facts

and we're going to do what's called a

fax relay switch over now with Cisco

facts relay we're going to do a fax

relay switchover and we're going to tell

the destination voice Gateway that we're

doing this fax relay switchover and it's

going to send us a codec acknowledgment

both voice gateways are going to do a

codec download

and essentially what's happening is that

we're going to change parameters like

before necessary to support facts so

terminal off vad playing with Jitter

buffers Echo cancellation Etc

so what's happening

is we're doing the codec download

we are then

I'm going to send an indication that the

codec download is done and get an

acknowledgment and the fax relay session

is established so the T30 analog fax

signals that we're getting from the psdn

are demodulated by the DSP on the

Gateway so that's the voice Gateway on

the left

packetized and sent Across The Voice

network as data and then the destination

voice Gateway

is going to remodulate those signals and

send them to the destination fax machine

as T30 so a little bit different

but um you know you can see certainly an

efficient way of doing things now this

is Cisco facts relay let me show you

t38 for both h323 and sip which again is

really really strikingly similar to

Cisco facts relay this is just a

standard we start with the t-38 or the

T30 session or voice call and T30 at the

other end again we wait for our said

tone and our disk message

with h323 we send a mode request from

the first voice gateway to the

terminating voice Gateway we get a mode

request acknowledgment

we then close the VoIP session and open

our t-38 channel so that's our data

stream

and then we have the 238 data stream and

we're done so again the originating

voice Gateway and then terminating voice

Gateway are

demodulating and remodulating

respectively

this signal for transport across the

network so guess what codec you need to

use on your network it doesn't freaking

matter anymore so this is the beauty if

you've got sites that are you know using

low bit rate codecs

like g729 you can still send fax across

those links

now with the Sip t-38 fax relay you're

going to see the same thing as you did

with h323 you're going to have the T30

signal Lane the voice call you know your

RTP stream between the endpoints you're

going to do the said tone in the disk

message but this time we're going to

send a sip invite indicating that we're

going to use 238 features within the sdp

message to the destination voice Gateway

or the terminating voice Gateway

it's going to send us back at 200 okay

we're going to then send an

acknowledgment

and nail up the t38 data streams between

the two voice gateways so pretty much

the the same kind of thing is three two

three now when we talk about configuring

fax relay I'm actually going to walk you

through an interactive tour here

we're not actually going to configure it

but I'm going to show you the flags that

you're going to throw this is so

trivially simple to set up that you're

going to have no problem with this so

standby while I bring up my putty

session and we will go into this fax

relay configuration

all right so we've got our putty session

here we'll go ahead and log into my pstn

Gateway and let me do a show run I

believe we've still got some

pass-through configuration stuff set

yeah so we've still got our modem Pastor

configured on this gate well we'll go

ahead and turn that off we'll say voice

service VoIP no modem pass through all

right so for t38 and Cisco facts relay

what we're going to do in same place

here fax protocol

and you've got options for Cisco or t38

if I say Cisco I hit enter and now I've

told it fax relay Cisco

pretty easy huh now if I go t-38 we have

a few more options we need to consider

here if we're doing t38 we're going to

say NSE to use Cisco's NSC to Signal the

t38 mode switch and then we've got

options for

um fallback if we can't negotiate t38

what would we fall back to

and then we've got options for

redundancies so let's go ahead and we'll

say fall back

and then we'll say pass through

gee 7-Eleven you law and then enter

other things I could do

is

um LS redundancy and I can set a

redundancy level and depending on the

redundancy level that I set

we're going to add additional data

information so that if a packet is lost

we can recreate the the data stream or

the fact Stream So basically this is a

kind of error correction so you can set

LS redundancy values to zero through

five

HS redundancy levels of zero through two

enter

yeah so show run we'll show you what

we've got going on here

so nothing crazy there now it appears

and I'm not sure why this is but it

appears that when you're doing the ls

and HS redundancy that you fall back to

Cisco pass facts pass through so I'm not

sure if that's a limitation or what you

know I'd kind of like to fall back to

g711 but uh again you know whatever

really that's all you're doing and like

I showed you before you can do this

either globally or at the dial peer

level I always try to do it globally

unless there's reasons that I can't

and I haven't found a reason yet that I

can't so

those are my standards but anyway you

know we've talked about what you're

going to need to know and understand to

work with facts within a Cisco voice

environment

you're gonna see it

um it's just the way it is please don't

have an aversion to facts don't run

screaming because you know oh I don't

want to set up facts because it's weird

I don't know how it works you know I

hear that too often and really it

doesn't have to be that way are there a

few uh Mysteries to it well there can be

until you get used to dealing with facts

on a network but once you understand the

ground rules it's not so bad so pass

through relay and t-37 storing forward

those are your options go ahead and

spend some time labbing this up for

yourself you know this is one of those

things that I'd recommend take it to the

next level go ahead and play with t-37

set up a mail server in a virtual

machine and you know actually route some

of these things I actually did that here

one of the notes actually that I'm going

to share with you for the t-37 stuff

that tripped me up a little bit

and it may be possible to solve but I

didn't spend the time on it because it's

not typical to what we'll be deploying

in the field but uh I had to use a PRI

interface on my router for the incoming

DNS I tried to use an fxs port and I had

problems um matching the outbound dial

pure lag and it probably had something

to do with Venus in the way that it's

detecting the destination I'm calling

but I didn't spend any time trying to

figure it out we just ran it on the PRI

but uh a lot of fun stuff here

and hopefully this will give you the

foundation you need to do facts in a

Cisco voice environment so with that I

want to say thanks for watching I know

this has been a long one I promise I try

to keep the others down I don't like

letting things run 30 minutes

um you know we try to to Target the 10

minute per video but this is one that I

think was well worth it and hopefully

you feel comfortable with the knowledge

of facts now that you didn't have before

so I'll see you guys in the next video

this pretty much wraps up this section

actually and we're going to get into

configuring call manager Express and

walk through building a basic CME

environment and we'll register some

phones and we'll create a dial plan and

you know it'll kind of be you know End

to End Soup To Nuts will work in CME

environment so I'll see you in the next

video we'll start working on call

manager Express or unified CME

foreign

[Music]

we're going to talk about Cisco unified

Communications manager Express or cucme

or CME as I'm going to call it CME is

something that you're going to find you

either like it a lot or you hate it a

lot personally I like CME I think that

it feels an interesting Niche and it

does what it does well

that said I don't think it is a

replacement for call manager but there

are certain environments where it uh you

know it just it fits really well so

let's Jump Right In and talk about what

is Cisco unified Communications manager

Express and we'll build some foundations

that we're going to use when we begin

configuring CME in the next video and

I'm actually going to walk you through a

configuration of the basic parameters

you need to understand about call

manager Express

CME is a or can be an all-in-one box

solution

and it's iOS based so that means this

runs on a router if you look at the

network diagram that I've got here I've

got you know an IP phone

a switch

what looks like a router but I'm calling

it a PBX slash voice Gateway and a pstn

connection this is our CME our call

manager Express so CME

began

as some something Cisco called its or IP

telephony services

or what we now refer to as telephony see

if I can spell here telephony service

v i c e within the router configuration

so you're going to see a lot of

references to telephony Services we go

through CMA

it supports as a as a conceptual product

or a device and I'm actually expanding a

little bit beyond the role of CME the

software to talk about the device the

device can be responsible for call

processing which is the CME or the

telephony services part of things

it can be your voice Gateway

it can be an iOS firewall if you know

the feature pack on the router that you

have supports that

obviously it's a router so it can handle

Communications for Wan connections and

Lan connections and routing between

subnets and all of the other stuff that

a router can do you know can be a DHCP

server Etc

with an additional module what we call a

cue or a Cisco Unity Express module it

can be a voicemail and an auto attendant

platform in fact the auto attendant

capabilities of cue are very cool the

same script editor and much of the same

script logic that we utilize in a larger

product called unified contact center

Express

is the same as what we use

within CME or within I should say within

cue so it's an all-in-one box solution

this is perfect for a single site

solution if I've got a branch office

with 20 or 30 phones

that are not frequently changing their

configuration you know I'm not always

adding or deleting phones or changing

directory numbers it's kind of a static

environment

CME is fantastic you can spin up a

router it can be that all-in-one device

for the site

and in fact you know I show a switch in

the diagram here but you get you guys

know that there are modules that can go

in the various models of Cisco routers

that are switch modules you know ether

switches Etc so we can you know

basically make this one box on-prem plug

your phones into it and walk away

so pretty cool uh interesting Niche

product now I said that it's you know it

fits the role of a single site solution

really well however you can interconnect

multiple sites I'm going to show you

that here on the next slide

this is one example of a multi-site

deployment for CME so if I look at each

location

each location would have IP phones and a

CMA and I'm assuming that we've got a

switch inside the router here so we've

got three sites

and they've each got to see me they've

got connectivity to the pstn which could

be an ISDN PRI they could be pots lines

you know fxo interfaces you know we've

got various models of IP phones you know

maybe there's you know 20 phones here

maybe there's five phones here maybe

there's 50 phones here

um one or more of these could have a

Cisco Unity Express module in it

and they're connected via a common ipwan

so

the CME at one location will be

responsible for the registration and

call processing of phones

at that location and will play no role

in the call processing of phones at

other locations so it's very much site

Centric

now I can

trunk these gateways together using

traditional means because after all they

are just voice gateways so I could build

for example an

h.323 trunk between the gateways simply

by using VoIP dial piers and allow the

phones within one location to call

phones within another location so it's

it's different than call manager where

you're going to put a device centrally

and manage it

um you know manage remote sites from it

and that we're going to distribute these

devices

um you know so that's a really great

textbook multi-site deployment example

you can also use a vpim for some

voicemail Internet working as well I'm

not going to get into vpm and in too

great a detail here but it is something

that's possible

when we talk about supported Hardware I

want to say this is not an exhaustive

list but it gives you a representative

sampling of the types of Hardware that

you can run Cisco CME on and CME is

versioned and you'll find that different

versions of Hardware with different dram

and Flash requirements can support

different ios's which then produce you

know version whatever of CME

looking at a small site you know a Cisco

1861

it's a fairly inexpensive router I can

support 15 phones with CME if I'm using

perhaps a 3825 you know I'm at 175

phones or I can really scale this thing

crazy huge in a 3945e with 450 phones

now

if I were designing a 250 or higher

endpoint system

personally I would probably be

persuading that customer to use call

manager

but you can do it with CME

supported endpoints a lot of the phones

that are supported with call manager are

also supported by CME and this is just a

brief list you know the 7900 series

phones the 99 51 70 ones

um you know a lot of the 6900 phones

they're all supported or I shouldn't say

they're all supported many of them are

supported by CME and you know check the

data sheet for the version of CME you're

deploying as this is not a complete list

finally I want to tell you that CME can

support both Sip and skinny devices

they are configured differently

whoops pushing the arrow the wrong

direction there when you're using skinny

devices we Define these as ephones and

their directory numbers or iPhone DMS

and when we're configuring sip devices

we're going to use voice register pools

but I'll get into more of that as we

actually go through an example of

configuring a CME I'm going to take a uh

2811 I think that I've got nearby

and Stage it up here in the next video

we're actually going to do a walk

through and configure basic CME

functions and features within that

device so for now I'm going to leave

this alone and say thanks for watching

this is your intro to CME and uh hang on

for the next video it's going to be a

lab and we'll walk you through base

platform configuration so for now good

luck with your studying and I'll see you

soon

[Music]

thank you

in this module we're going to go through

the exercise of configuring a basic

Cisco unified Communications manager

Express or call manager Express system

now we talked in the previous videos

about CME fundamentals and explained to

you the role that CME plays as a product

and the niches where it fits and I

wanted to walk through at least a a

reasonably typical installation of CME

just to show you exactly what kind of

configuration is is necessary and

possible with the product a lot of

Engineers that I've worked with in the

past have considered CME to be a bit of

a toy and I want to respond to that with

Ferrari is a bit of a toy as well so I

wouldn't necessarily

stereotype the product in a non-serious

way because it can really do some pretty

amazing things we're going to go ahead

and spend most of this video inside of a

terminal window and and do some

configuration of CME and walk through

you know what I I would say that a

typical config looks like and actually

register some phones and place some

calls so

this router that we're in this is one of

my 2911s that we've used before

and we've got it configured with an ISDN

PRI in fact it's pretty much serving the

role right now of just a traditional

voice Gateway you know if I do a show

dial peer voice summary you're going to

see that I've got a couple of Dial Dial

Piers Bill wanna one and 100 both are

pot style piers and they're pointing to

a trunk group that's got a PRI in it and

you know if I do a show ISDN status

you're going to see that you know the

pris and service we've got multiple

frame established Etc so you know it's a

normal voice Gateway right now we're

going to layer CME on top of this I'm

going to start out by walking you

through some of the basic CNE

configuration now there are two

different methods of configuring phones

in CME and I'm going to approach them

separately I'm going to walk you through

an all skinny example and then I'll come

back and teach you what you need to

understand about sip since really if

you're going to support skinny endpoints

and sip end points on a CME

it's almost like configuring two

completely separate systems you know I

don't think that that was the most

fantastic design that Cisco could come

up with but it is what it is and the

we'll walk through it

so it all starts with telephony services

in fact

um you'll see a a reference point from

time to time called its and its stands

for IP telephony service and it was

really the service functionality that

ultimately turned into the product

called manager express

telephony service is a menu access from

the config menu and there's a lot of

things we can do under televony service

relative to building a PBX you know CME

like I said you know it's a toy there's

a lot of really cool things you can do

with it

it can support auto registration of

devices or not I'm not going to show you

auto registration in this example you

can look that up on your own but we will

walk through some manual configuration

in fact the first thing I'm going to do

is say no Auto reg E phone so no Auto

red G phone under telephony Services is

going to disable that auto registration

now with CME you're going to Define

what's called an iPhone

and an iPhone DN now these are Concepts

relative to the skinny configuration you

can have a given number or a maximum

license number as well as a platform

maximum of iPhone's new phone DNS now I

said this is a 2911.

um my Max E phones on this platform are

42 you know as it's configured we're

going to go ahead and say Max E phones

10 for our configuration we can do a Max

DN of up to 144 on this platform I'm

going to go ahead and select 20. now now

that I've configured the Maxi phones and

Max at the ends I need to configure the

IP address where skinny phones are going

to register to and so this is a

combination this is the IP address and

the port like I said this is configured

like a typical router or a voice Gateway

so we've already got IP addresses on the

interfaces I'm going to say IP Source

address 10 10 210.1

port 2000.

I'm going to configure what we call a

system message system message is a CME

feature that lets you you know like it

says display a system message at the

bottom of the phone sometimes this will

be like the name of a business

um or you know your company name I'm

going to put

um for let's see here system message how

to network.com

seems like a good good example there

files that configure phones the XML

based configuration files just like on

call manager exists with CME because

obviously the phone doesn't know the

difference really from whether it's

registered to a CME or whether it's

registered to a call manner it's just a

skinny device you know speaking this

language you can store these config

files a couple of different places

in fact I'm going to show you CNF you

can store them location location either

to the default location which is called

system and I don't like to do that you

can store them to an external tftp

server which you know may make sense in

some environments but for me I'm going

to store them in Flash so we're going to

say CNF location Flash

now for each model phone

that I want to register to my CME I need

to define a load a phone load

we do this in call manager with Device

defaults we're going to do it in CME

with a command called load now what

we've got here and I'll paste two of

these in and then I'll walk you through

exactly what they're doing so we've got

a 7960 or 7940 image and the 7961 image

79 40s and 60s use the same firmware so

what we're doing here is we're saying

load we're specifying the model number

and then we're specifying the file name

without the file extension of the file

as it sits on our tftp server now what I

should tell you

is and we'll do it here in a moment I

haven't done it yet but you need to

upload

um a bunch of different files to the

router to enable full CME capabilities

now CME itself the feature set is just a

part of the router but there's a web

interface you can enable and I'm not

going to go into the web interface at

this point although we do talk about a

little bit in the CCNA voice videos

there is

um you know phone loads ringtones all

those kinds of things you'll have to put

them on the router you know they're a

download from cisco.com Etc so we'll

we'll show you what some of those look

like and how I've set my system up so

we've specified the loads we're going to

go ahead and Define a time zone Time

Dash Zone and depending where you're at

you can pick your time zone I am in

eastern time in the United States let me

find that that is 13. so we're going to

say time zone 13 for me lucky number 13.

now CME just like call manager can

support conference calls so we're going

to define the max number of conferences

that's between 1 and 16. I'm going to

just say four

now if you were doing web Administration

and had the GUI there's a command called

Web admin I'm not going to finish

setting it up but I want to show you

it's there whoops hang on web space

admin

and you would go in here and Define a

customer administrator assistant

administrator Etc like I said I'm not

playing with the GUI here so we're just

going to stick to the CLI stuff but I

did want you to understand that because

you'll try to set it up and you'll be

like why can't I authenticate I've

created a user account I want you to

know that there's a separate account

um call transfers are supported and you

can do a couple of different types of

call transfers let's go transfer method

we're going to go with a transfer system

and you can either do full blind full

consult or local consult I'm going to

configure mine as a full consult system

and the descriptions for the different

types of conferencing are right there on

your screen

finally I need to create my XML

configuration files and I'm going to do

that pretty much every time I go into

telephony services and make a change so

I'm going to say create CNF files and

hit enter and it's going to take a few

seconds to do that I mean I'm spinning

up XML configuration files like crazy

and once this is ready it'll give me a

prompt again we'll go through I'm going

to show you all the files I'm going to

show you how to do some tftp bindings

and I'm going to show you how to

configure the devices themselves and at

that point we'll be able to power some

switch ports up and register some phones

and make some calls

so it really is going to start with and

I can talk about this while the tftp or

I'm sorry while the uh the config files

are generating

um there's a bunch of files that you're

going to go on the tftp server on this

device I am using

um like I said before 79 61s and 79 40s

in this example I've actually got a

79.85 as well a video phone that I might

register up and show you

and

for you know any phone you put on your

CME just like on call manager you're

going to want to define a phone load now

I'm going to do it for the 41 and 60 I'm

sorry the 40 and 61. I'm not going to

mess with it for the 79.85 because it's

already running a compatible phone load

but uh I'm going to show you here that

we've got a whole bunch of different

commands and I'm going to paste them in

and then I'm going to walk you through

exactly what they're doing so that you

have a good understanding of how to use

them so we're going to control Z and I'm

going to go config T and I'm going to

paste a bunch of tftp server lines in

here and once they're pasted we're going

to do a show run pipe include tftp let

me show you what we've got going on here

so tftp server

flash colon slash phone slash 7940

slash p00308010.bin

Alias and then the file name again

what's Happening Here is I have chosen

to organize my flash instead of just

dumping everything in the root folder

hey if you want to dump everything in

the root folder have at it I like to be

a little more organized than that so I'm

basically defining my path and then

creating an alias within the tftp server

so that when the phone requests the file

the tftp server knows where to find it

so we've got these uh four files are

necessary for firmware for the 7940 all

of these files are necessary for the 79

61. so we've got all those tftp server

bindings in there now one more thing I'm

going to show you before we try to power

some phones up and register some

endpoints and place some calls and that

is ephones and iPhone DNS so in CMA

directory numbers have an object of

configuration that you need to do and

they're called iPhone DNS and they're

really simple to set up at a basic level

I'm going to drop four of them in right

now and just show you kind of what's

going on we've got iPhone dn1 dual line

and what dual line means is I've got one

button

that can be used but I can have two

calls on it now you can if you leave off

do line it's going to be a single which

is one call so you can't have call

waiting

um some versions of CME support

something called octoline Etc

that gives you eight

but we're going to do dual line on these

phones

now you'll see I pasted them in there

iPhone one dual line number 1000 E phone

dn2 dual line number two thousand then

you know three and four three thousand

to four thousand those are my extensions

lots of other things you can configure

on iPhone DNS they're a little beyond

the scope of the ccnp voice but you

should play with them in fact let me hit

a question mark and show you some of the

things you can do we've got call

forwarding parameters caller ID stuff um

we've got

um if Indian templates that you can

build and leverage

intercoms hunt stops for hunt groups

music on hold settings you know it's a

full-blown PBX so if you can think of

the feature it's probably in here you

know within reason so lots of things you

can do with the DN for now I'm just

gonna keep it at number and keep it

simple now that I've done iPhone DN I

need to configure whoops config team my

iPhone and I'm going to do this one at a

time and I'm going to show you what's

going on we're going to say ePhone

one

and then I'm going to say mac address

and I'm going to drop in the MAC address

of one of my 79 61s

and then I'm going to say type

79.61 in fact if I hit the question mark

after type this will give you an idea of

all the different model numbers

supported on the on the CME

now that I've done type I'm going to say

button one colon 1. now What's Happening

Here is when I say button one colon one

the first digit is the button on the

phone button number one

the second digit is the ePhone DN so

extension 1000 which was iPhone dn1 is

going to appear on the first button of

this phone

now just like I did that I'll paste in

another 79.61 we're going to call him

iPhone 2 with exactly the same settings

except he has button one colon two so

he'll have directory number two thousand

I'm going to paste two more devices in

one of them is a 7940 and the other is a

79.85 but again the exact same

configuration process is taking place on

all of these so we've created the phone

DNS we've created ephones in fact if I

do a show to left any service good show

command here we can see all of the

configuration parameters as they're

currently set up for to Lefty services

you can look at um

how qos tags are going to happen

we can see the loads I've specified the

Maxi phones Max DNS

any Locale settings conferencing dsps

srst you know pretty much everything

telephony service related and there's a

lot of other things you can look at

under show telephony service you can get

into such Lefty service tftp bindings or

iPhone DN or dial peer you just look at

lots of different types of things so

um you know play around in there a

little bit and see what's available to

you so we've configured to Lefty

Services we've configured the tftp

server we've created our iPhone DNS and

our earphones let me show you the file

system and we'll go ahead and register

some devices and actually I'm going to

go like that to the top level okay I'm

in the root of flash dir I've got a

folder called GUI and that's where my

CME GUI files are at I've got a music on

hold file in the root that's just where

I usually put it I've got a folder

called phone which contains all of my

firmwares and in some directories I've

got a folder called ringtones let me

show you ringtones

tones dir you know lots of different

wave files out there that are available

for ringtones and obviously they're all

specified in the ring list dot XML file

in fact if I want to show you that more

ring list.xml you can see you know the

ringtones that are going to be

advertised to my phone so I'm using the

default set here and really you know

it's set up it's ready to go let's go

ahead and get some phones on this here

so we're going to in fact let me show

you a little bit more about what's going

on the router and we'll register our

phones up we've got a DHCP pool set up

for phones you've done those before so

I'm not going to go over that in detail

option 150 is pointing to the IP address

of our call manager Express here

we've got our voice service VoIP

configured you know just like on a voice

Gateway and our t-38 you know not part

of this exercise here

um we've got voice class codec stuff

we've got a you know a T1 with six time

slots set up actually this is a PRI you

can see the PRI group there

um we've got all of our IP configuration

we've got um you know here's all those

tftp server definitions we put in

um you see our voice port for the PRI

you see the two dial Piers that I've got

configured and then the telephony

service stuff you know we've just done

that so I think you've got a good handle

on that our iPhone DMS and our earphones

and you know that's really the router

config so if I want to go ahead in fact

give me one moment while I connect into

a switch here my 3550 and bring some

ports into service

and we'll watch some phones come up and

register on the CME so one moment

interface fao2

no shutdown at f803 no shutdown and FAO

seven no shutdown

and end f809 no shut down all right so

I've brought some ports into service and

I'm going to show you command show E

phone

registered and it's going to show us

phones that have registered to our CME

now right now we've got nothing

um and you're going to start seeing some

messages of devices that register in

fact the first one there was my 79.85 it

actually registered so fast because it's

not a Poe device so it was already

powered on so we just had to wait a few

seconds for the spanning tree stuff to

happen we've got other phones coming up

it looks like the 7940 is in service

that's the second one you see there and

my 7961s are booting right now it won't

take them very long

um I want you to remember the iPhone

boot process again we covered this in

detail in the CCNA voice search exam

series but um you know the phone's

coming up it's requesting config files

or or I should say it's requesting the

itl file Stressless stuff and CTL files

um based on you know if it's been in a

security by default system before

it's going to then you know start asking

for config files and you know based on

the Mac address of the device and you

know we're going to give it ultimately

either a specific config file or a

generic XML default config file and it's

going to tell us what phone load or what

firmware load do you use and you know if

it's not what we have we're gonna reboot

and you know the next time we boot we're

going to ask for the files and fetch

them and download them and run a patch

and boot a couple times and finish that

process and uh you know then we'll be up

and in service these I've already done

that to so you you aren't going to have

to wait on that process to happen

there's one of the 7960 ones and the

other one will happen here momentarily

and then we'll go ahead and we'll make

some test calls and we'll make a call to

the pstn and that'll be it for skinny on

call manager Express

just about there

and it's registering now and it is

registered so that command I showed you

earlier showy phone registered now

you're gonna see some pretty nice output

there for the four phones that have

registered so you can see the Mac

addresses and it's registered in skinny

you're getting the IP address of the

device you see in the model number

you're seeing how the DNS are configured

on the buttons and the codecs and it's

pretty straightforward so four phones

are up let's make a test call debug ISDN

q931 I'm gonna pick up an IP phone

there's my speakerphone there I'm going

to press nine

and then five five five one two one two

we've matched a dial up here we've

placed an outbound call the calling

parties extension 1000 and you can hear

my test set ringing in the background

I'm going to go ahead and take it off

hook and answer the call and testing one

two three it's kind of kind of hard to

hear there because that's a that's a

piece of test equipment let me try this

here testing one two three one two three

so we've got a phone call up and you

know if I want to do you know show Voice

call summary you're going to see that

that call is connected and in a service

um oddly enough this is my my iPhone

so we can see I've got a g711 call up

and you know life is good everything's

working just the way it should I hung it

up there's our ISDN terminating the call

and we're done

so really straightforward really easy to

use you know I can do some phone to

phone calls right now you know we'll go

ahead and we'll call extension 3000.

you know there's a 79 61 calling a

79.40. you know I can answer it hello

hello hello you know I can press the

conference button and do a conference

call and and any of the other typical

PBX features that you'd be using

I'm gonna stop right here and push the

pause button I'm going to spin up some

config and show you the final piece of

CME that I want to make sure we go over

from a configuration standpoint and

we're going to set up a endpoint as a

SIP device using the voice register

Global and then we're going to wrap up

this video we're going to circle back

talk about troubleshooting in the next

video and that'll be it for CME so

standby one while I spin this up in

another window and I'll be right back

with you

all right we're ready to walk through

some SIP device configuration on CME it

took me a few more minutes actually took

me like an hour and a half

longer than I expected to get prepped up

and staged for this I ended up actually

um uploading some additional firmware I

realized that I didn't have any

available phones that already had a sip

image on them so I went ahead and took

one of the 7960 ones that I was using

and converted it from skinny to sip so

that's been taken care of it's out of

the way and things have been tested so

let's go ahead and deploy some

configuration necessary to run sip

phones on unified Communications manager

Express the first thing I want you to

understand is you're practically

configuring a completely different phone

system so where skinny phones you know

the core stuff is all about telephony

services and then the phones and DNS are

ephones and iPhone DNS well if you're

doing sip throw that all away because

it's completely different to set up

skinny endpoints and they can coexist

but to set up skinny endpoints on CME

you're going to go you to start out with

voice service VoIP

and we're going to say allow connections

sip to sip so obviously you know we're

going to have sip endpoints and then

perhaps they're going to hit some zip

dial Pierce you know maybe so we'll go

ahead and we'll set that up we're going

to say Sip and we're going to say

registrar server so we're going to be

the registrar server for these endpoints

and I'm going to say bind all Source

interface fast ethernet 0 0.30

so sip is turned on and we've set up our

register server and we've got it bound

to faoo.30 for media and signaling

we're going to type

voice register Global this is the

equivalent of telephony services so

voice register Global and we're going to

say mode

CME

Source address

10 10 to 10.1 Port 50 60. so again this

is our sip port

we're going to say load

79.61

term 61.default

we'll go ahead and do Max pool this is

the equivalent of the max e phone entry

so max pool 10 that'll be more than

enough and Max DN 10 Again more than

enough for the one device we're going to

set some localization settings up and

I'm going to say user Locale us and

network Locale us

we're going to set some time and date

stuff here everybody seems to want a

date format of 12 hour I'm sorry a time

format of 12 hour we'll do that a second

a date format where I'm at in the world

a month day year and in fact you can see

all the different options here day month

year and pretty much there's an option

for whatever uh you know whatever

they're doing in your neck of the world

so we've got the date format set I'm

going to say time format

um 12 hours or 24. I'm gonna go 12 hours

on mine and I'll say time zone I'm going

to be times in 12 but we'll show you

there are you know 56 different choices

so time zone 12 for me here in the

Eastern Time Zone zip phones unlike

skinny phones get their time directly

from an ntp server sip does not provide

time sync to them so we're going to

specify an ntp server for them to use

1010 210.1 I happen to be running an ntp

server on this router it's an NCP master

the next thing we're going to do is

Define our tftp path and some config

file stuff so we'll say tftp path

and then just like CME I'm sorry just

like skinny I'm spelling something wrong

here gftp oh I left the mode I was in

voice register global

that's what I meant to do there and

actually I made a mistake on the ntp

server I need a dash there there we go

that's why it exited the mode I I ended

up setting it for the router not for

voice register Global so um tftp path

and then I'll show you you've got flash

you can put it to a card you know a slot

or again tftp I'm going to say flash

colon we'll just let it live in the root

and I'm going to say file text this is

going to make my config files ASCII

format so that they're readable

and then the equivalent of create CNF

files for skinny for sip we're going to

say create profile and again I'm going

to do create profile a whole bunch so

we'll do it now just to show you how and

then we'll do it after we program our

phones to program a directory number

which is the equivalent of an iPhone DN

it's going to be a voice register DN you

know pick number one and we're going to

say number

6000. seems like a good choice and exit

now I'm going to define the equivalent

of the iPhone which we call a voice

register pool

voice register pool and I'm going to

call it one I'm going to do an ID and

you can either use the IP address Mac

address Etc I'm going to go ahead and

use my Mac address because it's

convenient for me if you have a static

IP on your phone you know maybe you'd

want to use that there's my Mac address

I'm going to say type 7961 and DTMF

relay RTP nte and you know you've got

I'll show you the options here you've

got sip notify sip kpml Cisco RTP but

I'm using RFC 2833 on my networks so RTP

nte it is

Okay so we've programmed a phone we've

programmed at the end

um I like to get in the habit of going

voice register global

and create profile

and hear momentarily we will be able to

boot our phone

and forget the error there

um we're going to be able to boot our

phone and get it registered and make a

call from it so let me do just that in

fact let me uh

turn a debug off that I had running and

I will tell this fund to reset and it'll

register with CME and we'll go ahead and

make a call so stand by one and once

this thing's up we'll make a test call

and that'll wrap up this video and we'll

move into the troubleshooting stuff

all right so I'm staring at the phone

and it didn't work

um in fact I'm not going to edit this

out I want to show you I missed

something fairly obvious I already know

what it is but uh I'm gonna walk you to

it here

and show you the mistake that I've made

you'll probably do this a bunch if

you're like I am

um let's see here whoops I blew right by

it here voice service VoIP okay look at

this

I created my phone and I created my DN

but I forgot to put the DN on the phone

so what I need to do is voice register

pool one and we're gonna say

hang on question mark here it is a

number

one

DN one so I'm basically tying the first

button to dn1 and uh voice register

global

create profile I'm going to tell the

phone to reboot again and that should

take care of my issue what it's telling

me is it's actually got a message on the

display and it says unprovisioned so it

was kind of a no-brainer it's like duh

you forgot to give it a directory number

so stand by we're gonna let the phone

Boot and uh we'll do that testing I

promise to

all right that phone is finished booting

we now have extension 6000 showing on it

and uh we're good to go we're ready to

make some test calls in fact I'm gonna

pick up a line here on a skinny phone at

three thousand we're gonna dial 6.000

and we've got a zip phone ringing I'll

answer it and you should hear the

familiar header dine so we've got calls

between phones

I'm gonna pick up the Sip end point and

dial

955-1212 and the test set across the

room is beeping at us as the PRI is

ringing and we'll go ahead and go off

hook and answer the call test one two

three you can hear my echo there a

little bit so that's working exactly the

way we want it to so we've got a couple

of things happening here actually we got

a lot of things happening on this router

at one time we've got um all of our

layer 3 stuff for you know routing we've

got telephony services for our skinny

phones we've got the voice register

services for our ZIP phones we've got a

PRI and that means it's all in one box

it's an amazing solution and that's how

you do call manager Express now there's

a GUI I'm not going to get into it as

part of this

um if you want to see it take a look at

the CCNA voice video I did knowing the

GUI is not really part of what Cisco

wants you to know for the C voice

content but you should have you know a

pretty decent Comfort level with

configuring call manager Express

with that I'm going to wrap up this

video and we're going to jump right into

the last section on CME and I'm going to

talk about a couple of different

troubleshooting commands and processes

that you need to think about and follow

when doing debugs on a CME when

something's not working quite right I've

certainly used plenty of them tonight as

I was creating this lab and with that

I'm going to shut up it's been a half

hour you guys need to take a break relax

get a cup of coffee come back and let's

do some troubleshooting on call manager

Express good luck with your studying and

I'll talk to you soon

[Music]

good

[Music]

module 19 troubleshooting CU CME so

we've talked about call manager Express

configuration and how to set up zip

phones and skinny phones and you saw

just a hint of troubleshooting but there

are a few troubleshooting processes I

want to make sure you're comfortable

with as you both prepare for the C voice

exam as well as prepare your skills for

being a voice engineer and supporting

these things

of the utmost importance before I talk

about anything else you have got to

understand the iPhone boot process this

applies to call manager call manager

Express and it is the thing you're going

to spend a lot of time doing you know

over the years as you support things so

let's start from the from the top here

the first thing you got to do is get

power to the device when you plug in a

Cisco phone to a Poe switch there are a

number of different uh you know things

that happen one of those is the switch

sensing that a Poe device has been

connected and applying power to the port

if you want some interesting reading go

read up on Fast link pulse it'll uh

it'll kind of open your eyes a little

bit to what's going on behind the scenes

once the fun power is on it's going to

load the stored firmware that's on it

and begin the boot process it's going to

identify via either CDP or lldp or you

know maybe you've programmed it into the

device what VLAN we're going to use for

voice traffic because that's the VLAN

we're going to send our DHCP request out

on so we're going to send out our DHCP

request we're going to obtain an IP

address and inside that DHCP request

we're going to receive something called

an option 150 address that is our boot

server

the boot server is the tftp server we're

going to request our configuration file

from the tftp server

we're going to retrieve either a

specific configuration file for us if

the you know call manager call manager

Express has one if it does not it's

going to send us a default configuration

file one of the values in this

configuration file which by the way is

an XML formatted document pull the thing

down and read it it's pretty cool we're

going to update the phone load it's

going to tell us what version of

firmware we need to be running now we

can upgrade or downgrade there are lots

of rules around you know where you can

get to and from and sign loads and

unsign loads and we're not going to get

into all of that and that's something

you can read on your own but you need to

understand we're going to upload I'm

sorry upgrade or downgrade the firmware

as necessary to match the system ideally

you want every phone in the system you

know of a model type to be running the

same phone load

we're then going to register with the

call server another parameter within

that configuration file is who is my

call manager actually it's what are

three call managers you know or up to

three who is my primary secondary and

tertiary so you're typically going to

see you know a subscriber you know

number one a subscriber number two or

maybe a publisher and then an srst

reference but uh so you're going to have

some capability for redundancy there

once you register with the call server

final parameters are going to be

configured via skinny things like what

is my directory number Etc so it's

really really critical you take the time

to understand this IP fund boot process

because it applies you know to a lot of

different types of troubleshooting that

you're going to be doing I'm going to

switch gears here for a moment and we're

going to jump into a console window

because most of this video is going to

be interactive and I think you're going

to learn a whole lot more from seeing

things happen in the router versus just

listening me talk and describe them so

here's my Cisco 2811 that we've been

using a lot and I want to just kind of

set the stage here I've got

um two Cisco 7961 phones on it I've got

a Cisco

7985 which is

um a video phone so those are all skinny

devices I've got a 7940 I should say

that I take that back one of the 7961 is

a skinny one of them is sick and I've

got a skinny 7940 so four amp points on

the on the device here

we're going to go through both Sip and

skinny troubleshooting processes with

you just to kind of scratch the surface

so one of the first things when you're

troubleshooting a Cisco phone is if you

know if it's not registering make sure

it's in the right VLAN so we're talking

about a switch where we um

where we program a voice VLAN you know

we're going to learn it either via CDP

or lldp make sure that's happening so

it's really more of a switching thing

not a router thing you can go to the

settings button on The Cisco iPhone and

dig in there and see what voice VLAN it

thinks it's using you want to verify the

IP addressing you know make sure you're

getting an address from the DHCP server

verify your tftp server make sure that

the DHCP server is sending you a boot

server this router that I'm on Show run

um pipe section ipdhcp I'm configuring a

DHCP pool of phones and you'll see in

that pool I've got an option 150 IP 1010

210 1. that is actually that's this

device but that is my tftp server when I

tell the device my tftp server it has

nothing to do with what call manager or

call manager Express I'm going to

register to that is only a tftp server

you're going to make sure that your um

TFT processes tftp process is working

debug t

FTP events is a great debug and in fact

if you really need to get granular you

can get into packets and you're going to

see what's happening with the boot

process you'll see the phone as it

requests the various configuration files

and things like file not found you know

those are gonna kind of clue you into

where your problem lies if you want to

verify endpoint registration for skinny

devices show

um let's see showy phone is a good place

to start show E phone a phone and

there's actually some options to this

you can go by types but I like

registered so showy phone registered it

shows me my three skinny devices I can

see the MAC address of each of the

devices I can see the state as

registered with skinny I can see the IP

address the model and I can see the

parameters for the buttons or what

directory numbers are assigned so that

is really really good we've got an

equivalent

on zip phones show voice register all

I'm sorry not I just spell things right

show voice register all and you can see

down here we've got voice register DN

and voice register pool so I can see the

directory numbers configured and I can

see the phones that are configured and

uh you know tells me a whole lot about

it here

um so that's great you know IP address

Etc

um if you need to run a debug for skinny

registration you know because it's not

working right debug iPhone

question mark and the one I like to use

is register you're gonna get a ton of

feedback on a phone in fact I'm going to

show you this one I like it so much and

I think it's so important that I really

want you to see what's going to happen

so we're going to power up a 79.40 here

and it's going to go ahead and register

and while while I'm waiting for that

thing to boot I'll talk a little bit

about what you're going to see in fact

one more command debug tftp events

um term mon you're gonna see the phone

come up and request some files it's

going to request this configuration file

um you know it's requesting the loads

file right now you can see that

um

that file tells it you know what

firmware to use Etc it's going to I'm

sorry the sep config tells it with

firmware the phone loads tells it what

files make up of the firmware it's happy

with everything it sees it's going ahead

and reaching out to different endpoints

and it's going to try to do a

registration so that's kind of the tftp

debug next we're going to see some

debugs for the actual registration

process itself I'm just kind of staring

at it here as it's going and you're

going to see it you know go through the

registration cycle

and it's really pretty straightforward

configuring IP okay

um

just waiting on it here I'm gonna

actually pause the video

no no I'm not no I'm not hang on the

phone actually rebooted I didn't expect

that to happen let's see what's going on

here

opening 10 10 203

let's see what we got going on here

and I'm actually going to let this roll

because this is the process we go

through unregister abnormally I wonder

why unregistered abnormally did I delete

it

let me see if I deleted it that is my 79

40 Show run

I like troubleshooting this is fun

um

we've got

see

we've got all of our tftp files

it's a laugh any service ah to left me

Services shut down there we go

television service no shut down make

sure to left any service is not shut

down

and actually it came up with the final I

found and then it's just now registered

there's also registration

um so that's the debug iPhone registered

yeah it kind of helps if you uh you turn

things on so lots of output here in fact

let me start at the beginning and show

you what we've got you know new skinny

socket accepted it tried to connect

here's the phone here's the MAC address

Etc

yeah blah blah blah

capability exchanges you know button

programming and it's done you know it's

up so

you know good debug to use to figure out

what's happening or not happening with a

skinny endpoint

um if you need to debug a sip

registration let me turn those debugs

off debug

um voice register events

it's going to be kind of similar let me

go ahead and reset a zip phone and let

you see what's happening with that it's

going to be just a little bit slower

because this is a 7961

but we're going to let the video keep

rolling because if I pause it I'm

probably going to miss it

um you know keep in mind

your life is going to be a lot easier if

you pick one type of phone to use on the

system you know sip or skinny you start

programming both and it's just you know

that much more complexity that you've

got to deal with

um you know when we start looking at

Best Practices for platform

configuration I know this is a

troubleshooting video but I'm going to

preach a little while we're waiting on

the phone to boot

um I like it when you run in your DHCP

server on an iOS device maybe it's a

voice Gateway maybe it's a router um I

just I'm a big fan of it I don't want to

do IP helpers and send stuff you know 14

hops down the network on different

subnets to some server that is going to

be maintenanced half the time anyway I

just I don't want to deal with that I

want core infrastructure providing DHCP

Services it's one less thing you have to

troubleshoot it's one less thing you

have to worry about you know monkeying

around with

use Smart configuration processes you

know make sure that your phone loads are

you know up to date or you know

standardized across the Enterprise make

sure it's worth working get into CME or

call manager and actually make sure that

the funds you're deploying are taking

the new loads don't just assume that

because you configured it it's working

actually check the phone make sure the

phone load that you think is getting

applied is getting applied because

you're going to have all kinds of wonky

Behavior with phones if you've got you

know 13 different or 20 different or

whatever firmware versions running on

other devices throughout your network

I'm staring at the zip phone it's

getting pretty close I think we're going

to see some output here pretty soon

and come on you can do it I wish these

things were faster they'll never be but

I wish they were like watching water

boil sometimes here we go updating

Locale that's the final step and here is

registering we should see some events

pop across the screen that's what we

were waiting for so we can see a

register request for 6000 from 10 10 204

contact matches pool one number list one

you can see it you know go through and

do the device provisioning so great way

to debug a SIP device registration and I

didn't mention it in the provisioning

videos but I'll say it now you can use I

never have

um so I can't tell you how well it's

going to work but you can use

third-party sip devices as long as

they're standards based with call

manager Express we can support digest

credentials and all those kinds of

things so give it a whirl try it play

with it and have fun with it

um

other things you want to understand

um with CNE everything's in Flash so

show Flash

you know it's going to give me some

information about all the files in the

various folders what they are whoops

sorry about that what they are where

they are do they exist

um you see this its folder these are all

those XML files the default ones anyway

that are being created for the various

Cisco phones so that's useful to know

if you want to do some show telephony

Services show telephony services

one of the things I like to do is check

the tftp bindings these are all the tftb

Bindings that telephony service is

making for the onboard tftp server on

the router you know config files Etc so

again if you're getting file not found

on tftp which that happens a lot

actually you know you can deal with it

um you know we talked about how to

verify some of the endpoint registration

with like showy phones so I think you're

good there

um you know anything else you know use

debugs use your log files

um you know and just kind of Step it one

thing at a time you know make sure I'm

gonna preach it again make sure you

understand that IP phone boot process it

is so crazy important most of your

troubleshooting is going to happen

with the registration cycle not after

the registration cycle so if you can get

that working you know you're 90 there

that's really all I needed to show you

in this video I told you it was going to

be kind of short although I'm about 15

minutes a little longer than I wanted

but you know you've got the fundamentals

you've got the foundations for CME this

is what they're going to be expecting

you to understand on the exam so go back

through the last three four videos that

I've done make sure you've got a good

grasp on what is CME where does it fit

how do I use it in my environment what

Earth capabilities what kind of Hardware

can it run on how do I program it and

how do I troubleshoot it and that's

going to be the ticket to your success

so with that I'm going to say thanks for

watching and we'll see you in the next

video we're going to start talking about

dial plans and how do I design a dial

plan you know ultimately we're going to

get into implementation of some dial

plans and then after that we're going to

talk about the unified border element or

Q which is a big topic a hot button

topic these days in the world of sip

we're living in so with that good

studying thanks for watching and I'll

see you in the next video

[Music]

foreign

[Music]

welcome to module 20. in this video

we're going to talk about numbering plan

Concepts and this is meant to be the

first of two pieces of a numbering plan

design and primer that we're walking

through so I want you to in this video

focus on understand what makes up a

numbering plan what are the things we

have to think about

what are the numbering plans that are

popular you know in the psdn today that

we're going to have to deal with on a

regular basis how do they work where do

they come from and then we'll segue into

in the next video some numbering plan

design considerations when you start

planning your own number of plans and

then ultimately we're going to go

through an exercise of implementing

number plans and we'll be dealing with

gateways and configuration and those

types of things but for this video we're

going to keep it you know pretty tight

from a conceptual perspective so first

off what exactly is a numbering plan and

you'll hear me refer to a numbering plan

as a dial plan from time to time

and really a numbering plan is a number

of things it's a scheme used to allocate

numbers

where ranges

be you know get assigned to Regions or

geographies or you know services like

cellular service in some areas you know

are assigned numbering plane ranges and

numbering plans Define the rules for a

number of allocation or assignment

so all of these things are a part of the

role of a numbering plan and you're

going to find that the international

world is using what we call the

e.164 numbering plan standard

and we'll go into a definition and

demonstrate how e164 works here actually

towards the tail end of this video

when we talk about the first numbering

plan you're most likely to run into

um and not just because I live in the

United States but uh because of its

popularity is the North American

numbering plan it was developed in 1947

and implemented a few years later and 19

countries in the North American area are

sharing the resources and using this

North American numbering plan it's

something like 800 area codes you've got

the United States Canada Bermuda Jamaica

the dominant Dominican Republic there's

you know plenty of others

but uh very very predominant in that

part of the world and here is an example

of a North American numbering plan

formatted number

you've got three div well first off it's

a 10 digit number the first three digits

are what we call the NPA

they can be with the first digit

starting from two through nine so any of

those digits and then two additional

digits so that's the NPA and then you've

got the NXX XXX

which is the local number portion of

things

which it can again start with two

through nine with remaining numbers

being X's so the first three numbers we

call that the area code

and the last seven numbers we call the

local number so this is what you're

going to run into in the United States

and other North American territories

very popular

widespread use

the North American numbering plan has

different reserved numbering codes and

and really styles of dialing

the first one I want to talk about is

called an ERC or an easy recognizable

code

um 888 are great examples of this and an

ERC is an annual number where the second

and third digits of the area code are

the same and they're just allocated for

special use purposes

we've got numbers called any two numbers

uh and Annie you've heard that term

before automatic number identification

service

um it's going to be used ultimately to

identify the type of originating station

and their two digit numbers or I should

say it's a pair of two digits we've got

cic codes carrier identification because

they're used for billing and call

routing and they are four digit numbers

within the North American number and

plan if and when you are placing an

international call it's always going to

begin with zero one one if you're

placing a long distance call it's always

going to begin with the one

in-state long distance or local calls

and traditionally this was long distance

but with

um

you know the numbering plan expansion

and growth Etc

we're starting to see local calls in

some areas require 10 digit dialing so

that's 10 digit codes

and seven digit numbers uh you know are

local calls so that's your traditional

nanp number and scheme some of the

reservations and the function of those

when you talk about special service

codes we've got eight of them two one

one three one one Etc all the way up

through 9-1-1 let's talk about what

these are and some of these are assigned

nationally and others within more local

geographies

2-1-1 is used for Community Information

in some areas 311 is used as a

non-emergency number to reach the police

or other government entities 4-1-1 is

local directory assistance 511 can be

used for local traffic information six

one one to contact repair service

711 can be used to contact a

telecommunications relay service 811 for

the business office and nine one one are

you familiar with that one I'm sure for

emergency telephone calls

another numbering plan or or you know

scheme that you're going to run into is

the etns the European telephony

numbering space and this exists in

parallel to the you know the Legacy or

the pre-existing geographically you know

related area codes and numbering plan

and we call this the 388 area code the

maximum European subscriber number

length is 15 digits and you're going to

find out that that pairs up rather well

with e164 standards and here's an

example of what one number might look

like so area code 388 with the access

code of 3 and you know the remaining

digits available to you

the etns has a couple of reserved

service types to Define you know how

numbers are assigned we've got

38831 for public service applications

3883 3 for customer service applications

3883.5 for corporate networks and 38837

for personal numbering

when you look at the numbering format

here it's made up of a country code or

group ID code plus a European service

code and those two those two sections

together are called the EFI the European

service identification

and finally the European subscriber

number

e.164 addressing is something you're

going to hear referred to a lot and an

e164 addressing is kind of the world we

live in it was developed by the itu

and it can be up to

15 digits in length so you remember me

mentioning in a previous slide about the

European numbers you know being 15

digits or up to 15 digits well an e164

number you know can also be up to 15

digits in length but it doesn't have to

be

each address in an e164 numbering plan

is globally unique and I mean that

literally

I sure hope we don't run out

um

because that would kind of suck and I

want you to see the format of an e164

number

like I said 15 digits in length the

first three digits

or can be the country code and I should

say up to three digits long

the next four digits are the National

destination code

uh digits 8 through 15 are the

subscriber code and when you put the

national destination code together with

the subscriber code

together they can occupy

um basically 15 digits minus the country

code so if the country code is one the

NSN can be 14 digits if the country code

is three digits long you know you're

you're losing another two so

um pretty much that sums up e164. you're

gonna see a lot of e164 numbers you know

through the course of this course in

this video series and telephony in

general

we're going to talk about things like

plus dialing and you know other types of

access so have a good understanding of

e164 numbers and how they play that

really sums it up for the uh you know

the fundamental concepts of numbering

plans in the next video we're going to

get into

actual numbering plan design and you

know how to create a scalable number and

plan and deal with numbering overlap or

prevent number and overlap and you know

ultimately you know how to route calls

based on these numbering plans so with

that we're going to wrap this one up I

want to say thanks for watching good

luck with your study and I'll see you in

the next video

[Music]

thank you

in this module we're going to talk about

some numbering plan design

considerations now this is only going to

take a couple of slides and a lot of

it's going to be obvious as you read it

and start to think about the impact on

your numbering planner dial plan design

on your environment these

recommendations are a combination of

things I've seen published things I've

seen other authors recommend things that

peers have done in the field and things

that I have done in the field and

lessons that I have learned and best

practices that I have developed over

time

there are no absolutes with numbering

plan design but there are a lot of

guidelines that you ought to follow and

the first of those guidelines is that a

numbering plan must be scalable I don't

care if you're designing a numbering

plan for a single site with 15 phones or

if you're designing a numbering plan for

150 000 phone deployment the same

planning thoughts need to go through

your head now you may not spend as much

time on them but you need to at least

think about them first off it's it's

critically important they use a

hierarchical design it is going to make

things so much easier

if there's a hierarchy to your numbering

plan as far as administrative burden

um as well as dealing with things from

the user side so you're going to have an

ease of provisioning new numbers or you

need to make sure that you can easily

provision new numbers within your

numbering plan you know so think about

how to allocate ranges for what you're

going to use now and what you're going

to use later you need to consider an

ease of routing you know make sure that

your patterns are created in such a way

that you know it's obvious how to get

where you're going you know one example

when I build a PBX or you know an IP

phone system my client is going to be

asked for a pstn access digit more often

than not they use the number nine but

I'm seeing others be popular you know

seven eight three and there's all kinds

of different options that you can use

but you know use that code or that PSD

and access digit to signify a call that

is leaving your premise it's going to

make it a lot easier in a voice Gateway

for it to make routing decisions if it

knows that everything that starts with

nine is automatically going out to the

psdn so

think about that think about numbers

summarization I want you to put numbers

in contiguous blocks

servicing

container you know similar geography so

if I've got an office in Columbus Ohio

and an office in the San Francisco

California I don't want some 5 000

extensions being in one site and some in

the other I want five thousands being a

one one thousands being the other you

know use some contiguous number in

schemes

consider scalability make sure you're

leaving space you know if I'm designing

a network for 5 000 endpoints I sure

better have like 10 or 15 or 20 000

numbering plan entries in my design

that's possible to accommodate for that

future growth you know what happens if I

make a stupid decision now well it's

going to be a pain in my butt later so

think about numbering plan and

extensions and how and where you

allocate things now not only for what

your immediate needs are but for you

know two three four five you know 600

growth it's not that hard to do

you want to make sure that you've got a

design that gives you ease of management

you know if everything's centrally

located obviously managing the numbering

plan is going to be a lot simpler so

we've talked about the hierarchy now

when planning specific

um you know specific guidelines minimize

the impact on your users don't do things

don't go into a site and change 5000

extensions on a whim that's going to

really cause chaos for your users and

create a lot of frustration and you're

not going to be the popular guy in the

IT department that week so try to

minimize the impact on the users and

likewise minimize the impact on the

existing environment if you've got

processes that have been in place for a

long time and they continue to meet your

business needs don't just change them on

a whim you know use what makes sense and

use some common sense within your design

create your dial plan or your numbering

plan to minimize the number of Digit

manipulations or translations that you

need to perform if if there's a way to

simplify translations by all means do it

and make sure you manage and plan for

growth if you don't plan for growth you

know it's going to just turn around and

bite you so make sure that you're doing

that as you work through your number in

plane design

dealing with overlapping numbers site

codes so I said try to avoid using

overlapping numbers it's not always

possible

um if you can't avoid overlapping

numbers let's talk about some strategies

you can Leverage

to deal with them

so site codes are one of those methods

when using site codes you're going to

use an enter site code

similar to a pstn access digit so I

mentioned before that a lot of times my

users will choose nine for an outside

line so when I'm programming my dial

plan you know I'll have patterns that

begin with nine

well

let's say that we're going to implement

site codes in our environment I may have

patterns that begin with a 7 or a six or

four or an eight or you know some other

digit and then we're going to have a

range of extensions behind that so

here's an example of what a site

code-based approach might look like so

in this example I wasn't able to have

non-overlapping ranges in fact

headquarters was the only site that

didn't have overlapping numbering ranges

um you know I've got extensions starting

with 1000 in headquarters and I've got

extensions that begin with 2000 and

branch location a b and c now this

didn't necessarily happen as a result of

a bad design perhaps I had some

Acquisitions and mergers in my company

or maybe we acquired a company and we

bought somebody and they had a bunch of

phones and pbxs and you know it just

didn't work out for us

so with this model we're going to

leverage the site code so for example I

might press eight

as a site code access digit and then I

would dial you know two one five four

and I'm sorry I skipped one I would

doubt eight I would dial my site code

and then I would dial two one five four

so eight two zero two one five four

would take me to branch location eight

or eight three zero two zero five four

would take me to branch location B so

site codes can definitely help you out

they're more of a traditional designer a

legacy design we're not necessarily or

at least I'm not encouraging people to

use them as much anymore I think there

are better ways of doing things

primarily using longer digit extensions

Etc

you know I mentioned that

um

it's not always possible to design a

non-overlapping numbering plan but if

you can do it you know you saw the

complexity in the last slide of site

codes I don't love them but they can't

solve some problems but if you're able

to design you know this is a

non-overlapping plan you're going to

have some Simplicity by having uniform

dialing you know make sure that your

users have a consistent experience if

I'm in site a

and I need to call site B

make sure that I'm using the same

process to call site b as someone in any

of my other 400 sites would use you know

I don't want a completely unique dial

plan and dialing behavior from every

location make it consistent your users

will thank you

select a single pattern length that you

can use you know either use four digit

extensions in your system everywhere or

use five digit extensions everywhere or

use 10 digit extensions everywhere or

use three I don't care pick a length and

stick with it do not get into a variable

length dial plan you will pull your hair

out it doesn't scale well your users

won't like it and it causes more

problems than benefits so use a single

pattern length four digits tend to just

Etc

here's an example of a non-overlapping

numbering plan again this is the ideal

design headquarters we have a we've

chosen a four digit example here so

we've got a four digit range of one XXX

so that's one thousand through one nine

nine nine branch location a has the 2000

Range and branch location B didn't need

quite as many numbers so we went from 3

100 to 32.99 and then we've got room for

future allocations of 3300 through eight

nine nine nine now you'll notice here I

didn't allocate numbers starting with

zero

and I didn't allocate numbers starting

with nine I like to leave those two

patterns out and leave myself some

flexibility for using those for other

things

here's another example of more along the

lines of what I'm recommending to my

clients these days this is a 10 digit

numbering plan I'm actually using the

full 10 digits of the North American

numbering plan to provision an extension

on a phone now

I'm going to create what I call

abbreviated dialing rules or

translations

that will allow a user to have a

simplified dialing experience so that

they don't have to dial all 10 digits to

reach the phones but the system knows

about it so my routing decisions are a

lot easier

so you know there's an example of what a

10 digit numbering plan might look like

expanded on that four-digit example

consider

bit number extension mapping you know

use dids where possible if

every phone in your Enterprise does not

have a did

but yet you still want to use 10 digit

numbering plan which I would recommend

create yourself some fictitious ranges

you know 555 is a perfect fictitious

range your users if unless you've got

your users dialing the full tension

number they don't need to know nor care

what the whole number is they just need

to know how to dial the people they're

trying to call so if you're using the

IDS try to match that to the extension

range and we'll show you an example of

doing that here my did range and

actually what I'm programming on the

phone is going to be 614 for

headquarters 5551 XXX and I'm going to

create abbreviated dialing rules so that

my users can simply dial one XXX now the

same thing would work

if I were using true four digit

extensions on the phone

and the pstn has its did you know the

614-55-1000 it's going to hit my system

I'm gonna look at the last four digits

in ring extension 1000 so by having

simplicity

and by and having the private extension

mapping match closely my pstn numbers

game it minimizes my dial plan

complexity

um is it possible to have you know a

whole bunch of weird directory numbers

and you know some other sequential did

range and map it to them yeah absolutely

and it's not a good idea it's going to

cause you a lot more trouble than it's

worth so try to avoid that

and finally tips for Optimum dial plan

design

first and foremost avoid ambiguity we

want to make sure that if you're dialing

a pattern that as much as possible that

is the only pattern you could possibly

match if we have to start making

decisions on what pattern was the best

and which one is the closest match your

dialing plan is going to be less than

ideal and your dialing experience is

going to be less than ideal avoid

overlap whenever possible it's not

always possible to have no overlap in

the system

but do as much as you can to get as

close to achieving that as you can and

you're going to be a lot happier

my personal recommendation if possible

use 10 digit or even full one E60 full

e164 extensions you know plus dialing

included if you want

four directory numbers within your PBX

it's going to make life simpler you know

if the carrier is

sending me

10 digit dinas on an inbound call

and I have 10 digit directory numbers

how many translation patterns do I have

to build in the system to Route calls

with phones absolutely none how many

dial plan entries do I have to build

absolutely none other than the obvious

existence of the directory number in the

system it's going to be like a connected

route if I'm talking about routing it's

just there because it exists and I'm

going to match it and I'm going to go to

it and things were simple

create translation patterns when you

need to use them in such a way as to

minimize dial plan entries if you can

have one pattern match a range of

numbers

that would be preferential to a hundred

patterns matching 100 unique numbers so

be smart about your translation patterns

use contiguous numbering ranges in a

geography so if headquarters can have

all extensions beginning with one and

Branch a extensions beginning with two

and Branch b extensions beginning with

three that is going to perform a lot

better for you and scale a lot better

than if you put you know even numbers in

one side and odd numbers in another I

mean it's goofy you could do it but it's

not going to make you any friends and

you're certainly not going to like

yourself a couple of years down the road

administering the system if you do that

consider your call coverage requirements

who can dial what you may find out that

patterns and and class restriction come

into your dial plan design and you may

want to make decisions based on that and

finally consider the numbering plan and

numbering type when you're making pstn

calls when we're going out of PRI

we've got a concept of numbering plan

and numbering type you know that's

something that you've heard me talk

about before and depending on what your

provider wants

they may want you to flag all calls as

unknown unknown

they may want you to flag long distance

calls as National unknown or national

ISDN or you know subscriber Nash you

know there's all kinds of different

variations on this ask your provider how

they want calls tagged because you can

influence that

a number of places and your dial plan

can come into play as part of how you

influence that so I know this was short

and sweet we ran about 15 minutes and I

think we've hit a lot of the high points

here on doing dial plan design and

considerations for creating a smart dial

plan in the next video and actually it's

probably going to span a couple of

videos we're going to talk about

implementing dial plans and we're going

to actually get into a Cisco router and

we're going to build dial Piers we're

going to start with a design and talk

about things like digit collection and

evaluation and how do we deal with

things like caller ID and how do we deal

with translation profiles and how do I

you know manipulate digits and how do I

strip digits or append digits and really

it's going to be programming dial plans

on gateways so with that we're going to

put the theory aside I'm going to say

thank you for studying with me I

appreciate you coming along for the ride

hopefully this has been helpful to you

and I'll see you in the next video

[Music]

thank you

foreign

welcome to module 22 and over the next

couple of videos we're going to be

talking about dial plan implementation

and what I mean by dial plan

implementation is programming the darn

stuff on the CME or on the Gateway

you're working on so we've talked about

dial plane design and how e164 works and

you've got a background on the North

American numbering plan and and really

you know a good primer on numbering in

general now let's talk about how

gateways or how CME both deal with

digits and understanding how to route

calls based on digit collection

so CME and voice gateways are going to

make the rounding decisions based on

dial peer matching and I've got two dial

Piers up here that I'm going to use for

an example dial pure voice 100 pots

which has a destination pattern of 555.

and dial pure voice 101 pots which has a

destination pattern of 555-1000.

how we match a Diop here and which Diop

here we select

is going to be based on how the digits

are collected

so the question I posed to you using

this example as a reference if a phone

dials 555-1000.

which style Pier will be matched and why

you might think that the answer is

obvious

but the answer is actually not so

obvious

obviously we want to get the best match

however

what gateways do what call manager does

depends largely on the device placing

the call and where those digits are

coming from so let's go through and

analyze how this process works

I want to talk about a concept here of

one digit at a time versus n block and

we're going to continue to reference

these same two dial peer examples

depending on what the sending device is

a CME or a Gateway can receive digits

either one at a time

or in Block which is all digits at once

skinny phones on CME are always going to

send one digit at a time and that's the

nature of the skinny protocol skinny is

a client server protocol and every time

you press a button on that skinny phone

it's sending a message to the call agent

or the call server you know whether

that's call manager call manager Express

and you're getting screen updates and

other types of feedback

and you know once the call agent has

collected all of your digits it's going

to make a call but it's dealing with it

one at a time

sip phones on the other hand can either

can send either using n block

so for example a sip invite message

contains the whole destination of a call

you're trying to make

or zip phones can also send digit by

digit if you're using kpml or the keypad

markup language and whether or not

you're using kpml or you're using End

Lock is going to be based on the type of

sip phone that you're using and really

it boils down to this old stuff

and newer stuff and we'll get into that

as we go

again looking at these two examples

which one will we match

it depends

looking at digit analysis on CME

let's look at the example you know that

we've continued to reference here and

talk about it in the context of sip

phones so there are two types of sip

phones there's sip type aphones which

are the early Cisco phones the 7940 and

7960.

and really they started off Life As

skinny phones and then sip code was

released so they could run that as well

and then there's the zip type B phones

which is pretty much everything else

since that point in time

if you look at this chart we show

endpoints of IP phones and gateways and

the protocols in use and we show you

that skinny is going to be a digit by

digit match

that's just the way skinny works

and a sip phone is going to either be

um

digit by digit if it's a type a phone

I'm sorry I'm saying that wrong

um it can be digit by digit

it can be m-block

or it can leverage sip dial rules so

digit by digit would be a type B phone

and my slide is actually wrong here I'm

going to have to correct that and kpml

that's true kpml is on type B phones but

you're going to be forced to use n block

you know if you're using a older phone

so here's what I want you to picture

like it's kind of like a cell phone if

you are typing the number

and then hit dial or hit send the

sending isn't actually happening until

you're hitting that dial button

so you can consider that when the digits

are being sent

that they're being sent all at once you

know upon pressing that dial button

if you take the phone off hook first

and then dial the digits you're likely

sending things digit by digit so with um

type 1 phones or type A phones you're

sending it in Block

with type B phones you're using kpml

which can be digit by digit or you can

leverage sip dial rules now on a Gateway

it's always going to be sent in Block so

think of the protocols you're using

um you know these aren't

aren't uh chatty little you know one

button at a time kind of thing so we're

going to evaluate the entire strings

so let's visualize the digit collection

process here and I'm going to use the

example number of 555 1000. if I'm a

user and I dial

555-1000 and I'm on a skinny phone

you'll see that the phone is sending

off hook messages and then the first

digit five and then the next two judges

five and then five

and then one zero zero zero and it's

doing it a message at a time

and the Gateway as returning or the CME

is returning the dial tone and updating

the screen is necessary to reflect what

you've done

with a sip type a phone which is going

to send thanks and block we're going to

send a message and really it's a sip

invite message and that message is going

to contain the entire string that we

want to dial

so five five five one thousand you know

we're sending a sip invite and then our

call progress messages are going to be

returned from the CME to the phone

and finally sip kpml digit by digit very

similar to skinny in that we're sending

it one digit at a time however call

progress messages just like in a zip

type a phone are going to be sent from

the CME back to your dialing device

so I think this chart or this

illustration really kind of sums it up

the answer really is it depends so back

to our question which diopir is going to

be matched

when we dial 555 1000

and again for reference here are our two

dial pairs dial pure voice 100 pots with

a destination pattern of 555 and dial

pure voice 101 plots with the

destination pattern of five five five

one thousand

the answer is still it depends

so if it's a skinny phone

in this example because we're doing it a

digit at a time we're going to have

determined that we found a match

on dialpear 100

once we parse that third digit because

we'll go five five five I gotta match

if we're using Sip n block

we're going to have an exact match on

dialp 101 after all of the digits have

been received and we analyzed based on

the string of all digits

so as you'll see

there's you know some pretty

um

obvious things happening

to do digit selection or do to do route

selection for voice calls and it's not

always cut and dry how it's happening

the lesson to learn here and what you

should take away from this for prepping

for your C voice exam is that different

endpoints behave differently and you

really need to understand the specifics

of your environment to predict the

behavior

now in the next couple of videos we're

going to talk about actually deploying

the dial plane onto the gateways we're

going to get into digit manipulation so

digit stripping and and managing the

number of digits we forward and some of

the various tests you can leverage and

translation rules that we can do and

really kind of Soup To Nuts typical

voice configuration on an iOS device so

with that I'm going to say thanks for

watching and good luck with your study

and I'll see you in the next video

[Music]

thank you

in this module we're going to talk about

digit manipulation on the routers now in

the last video we talked about or

actually the last couple videos we've

talked about dial plan and numbering

plan design and the kind of

considerations you need to make when

designing your dial plan like avoiding

overlapping numbers and using contiguous

blocks of numbers for certain

geographies and we've talked about you

know how to come up with a good design

but we really haven't really you know

shown you where the rubber meets the

road where we're dealing with uh you

know performing all these kinds of

actions and that's happening with

invoice gateways and within your your

call agents so let's take a video or two

I'm not exactly sure how far I'm going

to dig into this one I want to kind of

keep the time manageable

but let's talk about digit manipulation

and what I want you to understand is in

every voice environment you're going to

have a scenario where you're dealing

with digit manipulation so we need to

understand the best ways to deal with

manipulating digits when you need to do

it so we're talking about things like

adding or appending or pre-pending or

stripping or you know doing all these

types of things I've got

a session open here to our 2811 router

our headquarters router and I'm going to

walk you through a number of different

types of types of Digit manipulation

scenarios that are common

when dealing with voice gateways or CME

environments now let me give you just a

quick primer of the system this is the

CME we were working on previously and we

have just so that you're familiar with

what's going on we've configured a trunk

group for our pstn facing pris now

there's only one here but if we had more

than one we could simply uh you know add

them to the trunk group

and I'm doing that as a a method to

minimize my dial pairs so instead of

pointing dial pairs to Ports I'm going

to point dial Piers to trunk groups and

trunk groups can contain multiple pris

we've got Sip and skinny phones

registered on CME you know so here's the

voice register Global and the ends and

you'll see the pools below that's all of

the Sip devices we've got as we keep

going here you know we've got our T1

controller up into service and I've got

six time slots in service for our PRI

we've got let me keep going here here's

some dial Piers you know telephony

services for our skinny phones couple of

numbers

and if we do Show e-phone registered

whoops registered you can see we've got

a couple of devices up and in service

um show ISDN status

is going to show us multiple frame

established and I've got one of my test

sets conveniently connected to the

Gateway and we're able to you know make

some test calls with it so if I want to

let me turn it up here

so we can actually see what's going on

set up backlight all right I can read it

now if I want to dial for example in

fact let me turn on a debug debug ISDN

q931 term mon let's say I just dial it

um I'm going to call extension 2000.

all right so we're going to place it in

my call

let's see dial two thousand two oh oh

on hook and you can hear the phone

ringing in the background now as I look

I'll go ahead and end the call here as I

look at the ISDN debug here I can show

you that the call coming in

is destined to a four digit number

and because my phone this is what we

call for did you name this because my

phone has a four digit extension then it

matches the dial peer entry for the

phone because remember in CME show dial

peer voice summary

I'm going to have

dial piers

that relate to the phone so you know I'm

looking at the route planner or the

looking at the dial plan and I match it

and I ring the phone

let's say

I didn't have four digit extensions

maybe I had five digit extensions and my

phone was one two thousand

I would need to be able to up to prefix

or to append a number

two

make the two thousand that's coming in

look like one two thousand or maybe I

have

um

let's see here maybe I have four digit

dinus and three digit extensions maybe

my extension is 500 but calls are coming

in to twenty five hundred then I would

need to do some digit stripping

or perhaps

um you know a call coming in to 2000

really needs to Ring to five thousand so

I'm gonna have some more advanced

modifications to do so let's walk

through you know some of the simple

stuff here that we can deal with

and you know keep in mind the same

thing's true in the outbound Direction

in fact let's look at outbound first

Show run pipe again dial up here

if we look at the dial Piers you're

going to see

the pot style up here that I have here

dial up your voice 100 Bots is using a

destination pattern of nine and seven

digits and I say forward digits seven

what I'm doing here is I'm actually

telling the dial up here

to only forward the last seven digits so

don't forward my PSD and access code of

9. so I'm effectively stripping the nine

um I could also use a command called

digit strip on a dial pair

and it's going to strip the Matched

digits in the destination string so you

know that comes in handy for like wild

card kind of stuff

so you can do digit manipulation that

way

we can do numbering expansion where we

take a small number and turn it into

something else you know an inflate or

deflate thing so we can use voice

translation profiles and rules Etc

so you know pretty pretty

straightforward here but lots of things

you can do

lots and lots and lots of things you can

do so let's see here

um we're already showing you an example

of forward digits let's let's actually

go through that one forward digit seven

so what I'm going to do dial up your

voice 100 Bots I'm going to change it in

fact let me make a call from it and then

I'll show you what happens when I change

it so I'm going to Dial 9

55-1212

I'm ringing outbound

with a called party number of five five

one two one two because I did just what

I said I was going to do I um I strip

that nine let's change the behavior so

that I don't

so let's look at the dial up here again

uh 100 so config T die up here voice

100 pots and we'll say forward digits

all so now I'm not stripping now watch

what happens on my debug I'm gonna dial

nine five five five one two one two and

now the out Mount call is ringing with

that nine still attached

Maybe

I needed to only send four digits out to

the pstn that's kind of unusual but

let's say I wanted to do four digits

four again on that same dial up here

I'll place the same call nine five five

five one two one two and you're gonna

see that now I'm simply calling the call

party of one two one two so it's very

easy to deal with stripping digits

now let's talk about appending digits

we'll go back to that dial pair and

we'll say forward digits

um let's see I'm going to say four

digits seven

so it's a normal seven digit call now

let's say that for whatever reason I

needed to dial a prefix code

maybe it's because this is a tie line to

another PBX so I need to leverage that

style plan or whatever but let's say I

need to do append a five

all right now I'm going to pick

something a little easier to see a 2 to

the beginning of all these numbers I

could say again on that same Diop here

um

that I could do let's see here

thinking about the best way to do that

I could do prefix

let's see here

yeah I think prefix is what I'm going to

do here prefix question mark and I'm

just going to say two

so again show run pipe begin dial

pure voice you'll see that 100 that I've

got here so our destination pattern

starts with a 9 and the seven digits

long

we're going to forward seven digits so

we're going to drop the nine and we're

going to prefix the two so we should see

a two

five five five one two one two coming up

to the pstn so we're gonna say nine five

five five one two one two and we should

send down the ISDN a two five five five

one two one two which we're doing so

right there we're showing you the basics

how do I strip digits

and how do I

manipulate or prefix now

if I had a a Diop here

that was perhaps variable length

and wild card based let's see here let's

do dial pure voice we'll just call it

150 pots and let's see how the other

one's configured here

we're using

trunk group PRI all right we'll point it

to the same trunk group

and then we'll say destination pattern

nine

zero one one t that's the perfect

example for an outbound international

call now what's going to happen here is

I'm by default going to strip the

9011 because that was the exact

portion of the match

I'm not doing any kind of forward digits

thing here so let me show you what's

going to happen so I'm going to go off

hook I'm going to say nine zero one one

and I'm going to dial a variable length

number five six seven eight nine yeah I

think I hit the buttons there and what

you're going to see happening is okay so

I skipped to seven so five six eight

nine is what I dialed so because

nine zero one one

was the specific

entry in my Diop here it matched it and

stripped it automatically

so what I'm going to have to do to make

that call successful

is prefix

prefix zero one one

so now when I call

nine zero one one five six seven eight

nine

you're gonna see zero one one

five six seven eight nine so I think

we've kind of covered prefixing and

stripping adequately the next thing you

need to understand how to do is number

expansion

um

and what you can do with number

expansion is you can turn

one type of digits into another type of

digits so for example let's say num

expansion and we'll say

[Music]

um

five five five

what's a good extension I've got a 2000

number here we use that two dot dot dot

two six one four five five five two dot

dot dot now check this out when I make a

call

and it might look goofy in the debug and

I might even not match it up here

because I haven't looked but I'm just

going to dial the digits and then roll

with it and fix what breaks but I'm

gonna go ahead and

go off hook and we're gonna say nine

five five five five two oh oh

so I've dialed 555 2000 oh actually I

matched the wrong pattern there that's

not going to be terribly useful for me

let me

change some things up here Show run pipe

again dial pure voice

whoops I scrolled right by it

101

150 let's go back to that 100

dial

pure voice 100 pots we'll say no prefix

two all right so four digit seven so

that's back to looking normal a number

expansion actually you know I'm going to

use it as an inbound test instead of an

outbound test I think it's going to work

better so what I'm going to dial in fact

let me do Show run pipe include num

we're gonna see the number expansion so

5552 dot dot dot is going to try to rank

614-552 dot dot dot so let's go ahead to

that E phone

um let's see here Show run and I'm going

to modify the number so the phone

actually has a 10 digit number on it so

we're going to present four digits from

the psdn expand it into the 10 digits

and ring the phone so where's it at

there's iPhone DN I'm gonna use the

phone with extension and 2 000. so you

found the N2 okay so we're going to

config T say e phone

DN to do a line and I'm going to say

number

six one four five five five two thousand

so now it's got a a full

10 digit number on it we're going to go

ahead and reset that phone so we'll say

config T we'll say e phone

two

restart

and that will cause it to re-register

and I'm staring at it now

and now it has a 10 digit number on it

perfect so if I do show E phone Reg

you're going to see iPhone 2

has a 10 digit DN on it so call is going

to come in from the psdn to 2000. should

hit number expansion and again let me

show you a show run

pipe include num expansion we're going

to take that 5552 dot dot dot

and expand it to six one four five five

five two dot dot dot so I'm going to

call

let's see here off hook

whoops hang on

getting ahead of myself here say

5552000 dial

so what happened is I called in to

555-2000.

we matched number expansion and we

expanded five five five two thousand

into six one four five five five two

thousand and hello hello we can ring the

phone answer it and hang it up and we're

good to go so that shows you the first

couple of strategies relative to

manipulating the dialed number

we've talked about digit stripping we've

talked about prefixing and we've talked

about number expansion I'm going to stop

this video here and just to kind of keep

things short and we're going to go into

some caller ID manipulation or calling

party manipulation and then we're going

to get into the big boy the voice

translation profile or how do I make the

real magic happen when I want to change

something crazy weird into something

else crazy weird so for this video we're

going to say thanks for watching I'll

see you in the next video where we talk

about additional

a digit manipulation techniques and uh

thanks for viewing I'll see you soon

[Music]

welcome to part two of our digit

manipulation video series and it seems

like in any way and we talked in the

last video about some of the the

fundamental digit manipulation Concepts

that you're going to deal with on a CME

or a voice Gateway and you know

prefixing digits and messing with the

forward digits so that you can

effectively strip digits when placing

calls and really we were focusing on

basic manipulation of called party

information we were worried about the

destination now there's one more bit of

information you need to worry about and

that's the calling party so we're going

to go ahead and jump into a router here

this is a different Gateway this is

actually my my pstn Gateway that I use

for the lab here and we're gonna I'm

gonna bring your attention to some

configuration relative to for example an

fxs port

so I'm showing you here voiceport one

so one one one that should be my fxs

port Let's uh let's see if that's right

uh show Voice or is it voice Dash port

or just voice port

that's when I want to show voiceport

summary and we are fxs so one one one

is a loop start fxs you know plain old

analog phone so nothing crazy there if I

want to go ahead and take it off hook

here and we'll do it again you can see

that it is actually 110 is 101 fxs

um yeah showed it go off hook so one one

zeros one we'll focus on now right now

if I call something

and uh let's see here I'm gonna check

and see if

um I've got anything dialable on this

let's do a show dial here voice summary

I had some psdn patterns

oh where do I want to send it 10 11. uh

give me a second I'm going to go ahead

and hook up that fxo line and we'll

actually send some calls to the pstn I

want to show you some caller ID features

so standby one while I connect the fxo

port all right we've got the fxo port in

service on a pstn connection available

here

um since the pots line we're going to

have to run a slightly different debug

but what I want to show you is debug

um voice CC API and out

term on whoops term come on I'm going to

make a call to a toll-free number here

so we're going to go off hook and you'll

see some some debugs here and I'm going

to say nine

1-800-444-44444

now as I place the call

I've got

obviously a called party number but I

don't have any calling party information

and I'm going to hang this up and then

we'll scroll through the debugs and I'll

I'll find where to show you that at

see here

where is it

takes me a minute I don't do these

debugs all that often

and usually I'm searching here we go so

I've got a called number

but I don't have a calling number

it's blank I don't have any caller ID

on that voiceport

if I want to set up caller ID on that

voice Port so that this actually

presents some caller ID to people

what I'll do is I'll actually go to that

voice Port what did I say that was one

zero zero

show voice Port summary

one one zero we'll say voice

Port one one zero we're going to say

station

ID and you can do name

or number I'm going to go ahead and say

number

and I'm going to say six one four five

five five one thousand

and we're going to use that as the

station ID number

and I'm going to go ahead and make a

call

and now I should send that caller ID 9

1-800-44-44444

and there we route

and I will go ahead and hang it up and I

actually saw it fly across the screen

we'll scroll back and look at it and

it's right here calling number equals

six one four five five five one thousand

and obviously we're doing that same

800 number destination so

manipulating caller ID on things like

fxo ports It's relatively

straightforward as is on a PRI you know

you can you know send whatever type of

called number I'm sorry or calling the

number you really want sometimes a

carrier will um

not allow calls originating from

incorrect numbers

other times they will

and I'm really not going to get too deep

into the calling party manipulation but

I did want you to understand that those

are configurable parameters and there

are things you can do with them but the

real topic for this video the thing that

I'm I'm gonna dive crazy deep into in

fact let me open a different terminal

window to do it because we're going to

hop in a different Gateway is regular

expressions and understanding what you

can do with voice translation profiles

so let me log in to the other device

here and bring it down into the window

this is the same router we were using

earlier our HQ router

and

voice translation rules and translation

profiles work together

you can attach a voice translation

profile to lots of different things you

can attach it to voice dial peers voice

ports Trump groups you know lots of

different ways you can deal with them

but I want you to think of voice

translation profiles and voice

translation rules as ways to turn

anything you want

into just about anything else you want

I use them a lot on inbound calls

um

you know I'll I'll have a a did that was

diode that I want to redirect to another

destination

and you know so I can do that now I want

to show you

how to configure voice translation rules

and how to map them I'm going to use a

voice interface or a voice Port as my

example but keep in mind these can be

applied lots of different places lots of

different ways but once you've once

you've seen the basics it's pretty

straightforward so the first thing I'm

going to do is let me think about the

scenario I want to map out I'm going to

accept an incoming call

to a

10 digit number

six one four five five five one two one

two

and I'm going to redirect it

to extension

3 000 okay

so we're going to go config t

and I'm going to create a voice

translation rule so voice translation

rule one

enter

and I'm going to define rule

one

and you're going to define a matching

pattern and the matching pattern is

going to be contained within these

forward slashes

and then we're going to Define

a substitution pattern what to put in it

and it's going to be in a pair of

forward slashes so I said I'm going to

and this is regular expression stuff so

you'll want to see how Cisco uses regex

I'm going to use a carrot which means if

the number begins with

six one four five five five one two one

two

then turn it into three thousand

I could technically do this

and do an exact match but I like using

begins with and my my rules most of the

time there are lots of different regular

expression characters you can use you

know there's a dollar sign which is used

for matching Expressions at the end of a

line you can have

um I mean it's just gobs and gobs of

different kinds of regular expression

variables and wild cards and things so

I'll leave that as an exercise for you

to research but this is the basic

structure of a translation Rule now that

I've created the translation rule

I'm going to

create a voice translation profile and

the profile is going to contain the rule

and then I'm going to apply the profile

somewhere so we're going to say exit and

I'm going to say voice

translation profile and we'll just call

it inbound

because remember I'm going to manipulate

inbound digits and I'm going to say

translate called

one so the one is going to revert back

to this voice translation rule one that

we had

now so I've got a profile and I've tied

it

to a rule

now I'm going to put that

on a voice port for inbound calls so

let's take a look and see what we've got

here show usdn status

uh serial zero zero zero so we're gonna

go voice Port zero zero zero colon 23.

and I'm gonna say now that I've got it

on the voiceport

oh let's see here

let me go translation profile

and then we're going to say incoming or

outgoing so I'm going to say incoming

and then I'm going to give it the

profile name which was inbound I think

is what I called it yeah I called it

inbound and I'm going to hit enter

so now what should happen

and if it doesn't I'm gonna not cut it

out and we'll troubleshoot it together

here in real time

but what should happen is I'll make a

call across the PRI

uh coming in from the pstn I'm going to

present

614-555-1212 and what should happen is I

should match it I'm going to rewrite the

destination to 3 000 and I should ring a

phone so let's see if that's what

happens in fact let's see if I've got

any debugs on I do I like q931 we'll go

ahead and leave those the way they are

so I'm going to say six one four five

five five one two one two and I'm going

to hit the hook and dial and

whoops term on would be helpful so you

actually missed the debug you can

probably hear the phone ringing behind

me in the background I'm going to end

the call and we'll we'll place it again

with a slightly better debug output so

here we go again six one four five five

five one two one two send so extension

3000 is now ringing and you'll see I

dialed the call to party was

614-55-1212 but extension 3000 is

ringing because we leveraged that voice

translation profile so this is a good

example of how you might match a did

number

that is not a pattern match

to or you know not an obvious pattern

match anyway to the destination device

you want to ring

you can also manipulate the calling

party or the caller ID information like

I showed you before but for me primarily

I'm messing with the cold party

information

so we've applied it to the port now

there's you know I really made a simple

rule that was kind of a get your feet

wet kind of rule there's so many things

you can do

with input and output strings

um you know let's let's talk about one

some of them might look like so we'll go

again to voice translation

rule one and then we'll say rule

two we'll create another rule under it

and we'll say match anything that begins

with a nine

and turn it into nothing so basically

we're going to do a digit strip

so that's a pretty you know pretty

straightforward way to do things in fact

let's make a call from the pstn I'm

going to dial not whoops that's not what

I wanted to hit I'm going to Dial 9

3 000. we're going to strip the nine and

it's going to ring at 3 000. so we're

gonna say nine three oh oh dial

whoops I didn't have uh quite enough

digits there that was 9 300 so 9

300 dial so we're stripping the nine

that I've got coming in and we're

ringing three thousand so again another

very simple example let me give you a

show run just so you can see what this

translation uh is you know structure is

taking on so here it is voice

translation rule one

so I've got two rules that we've put in

there so far and we're dealing with

these at the inbound voice Port so let's

let's create some more complex rules

translation rule one and we'll say rule

three begin or you know and we're gonna

match

um anything that starts with

um

let's see here

a three four or a five

dot dot dot

and we're going to convert it to three

thousand

all right so if I dial three thousand

three thousand one Thirty ninety nine

four thousand five thousand any of those

numbers we're going to turn into a three

thousand so let's see if it works so I'm

gonna dial a

five oops hang on ignore that I hit send

too fast I'm gonna dial 5000.

and it's gonna yeah come on I'm missing

digits here get some sticky keys on my

test set five zero zero zero send five

thousand is ringing in I'm matching that

translation Rule and ringing extension

three thousand if I want to dial 4206

again I'm matching that same Rule and

continuing to ring extension three

thousand so lots of really really cool

things you can do with these voice

translation rules let me show you a

couple more examples voice translation

rule one rule four

uh let's say that we want to um

oh I don't know let's rewrite a little

complex pattern

so rule four if it starts

actually not starts we're gonna say

special character

okay nine backslash and I'm reading this

out of an example in a textbook so keep

in mind you know you have to spend some

time and and create your own scratch Pad

of a regex to use

so it starts with a nine

or or has a nine special character

parentheses and then you'll see that I'm

putting these Escape characters between

the parentheses

um and then we're gonna say

boom

begins with one zero

and then forget me saying begins with

I'm just kind of talking to myself as I

think through this

dot star

backslash parenthesis

all right so there's our match pattern

and then

we're going to turn it into

forward slash

backslash

one

six one four

two forward slash I screwed it up there

was no way I was going to get that

perfect the first try so we have

parentheses around the nines we've got

brackets here and we're going to need a

parenthesis

before the bracket

there we go

much better now

what's going to happen is if I call nine

five five five one thousand

it should convert it to nine

one six one four five five five one

thousand if I wrote the rule right like

I said I was referencing another example

here

but what I'm going to actually do is

we'll put that number

on a phone so config t e phone dn2

number

this should be nine one six one four

five five

um

what do they say one thousand

all right now let's see if it actually

works so show run

and here are the rules we have

and actually I've got to reset my iPhone

iPhone 2 restart

and the rule I want to try to match is

this last one here this rule four

so I'm staring at the phone

and it looks like it's back up

so I'm gonna call

955100 send

unassigned number it did not like me

nine five five five one thousand let's

look at that rule and see what I'm doing

wrong we'll pick it apart a little bit

here

some of the fun of trying to do these

things on the fly so there we go we've

got on our clipboard go ahead and throw

it into a notepad here and let's uh

let's work the logic out so

if it begins with a so first digit would

be a nine

and then

a begins with one or zero so one

then a DOT star whoops dot star so a DOT

is going to mean match any single

character

and Then star is going to repeat the

previous regex zero or more times

so that's like a four digit pattern so

far or more so nine one something and

then whatever else as a length so that's

our match so for example if it was 9

1 6 and then now we're to Star

um actually nine one

and then

555100 so nine one five and then 5100 so

that's where we are so far

and we're gonna take that

and we're gonna say

um

turn it into

let's see here

the rule we have is a

backslash one six one four a backslash

is a special meaning of the next

character

so slash one so that has to do with

variable pulling from the beginning okay

and then six one four

slash two

Okay so

nine pulls over

one six one four

and then

that should be

five five five one thousand

but clearly something's not working

there so let's do this you know what I

think it makes sense on paper let's see

what's actually being translated and

we'll uh we'll pick it apart from there

there's a way to do a test

on a pattern and on a voice translation

profile

and it'll make it a little bit easier to

diagnose what's going on

so let's do a test on that let's say

test voice translation rule

the number is going to be one we're

going to type the input number

which is going to be what did I say I

was going to dial I was going to Dial 9

555 1000. enter

matched with rule two two

and it translated it to five five one

thousand well wait a minute rule two

aha I have some overlapping rules let's

get rid of rule two

voice translation

rule one no rule two

and this is exactly the kind of stuff

that I run into you know all the time

when dealing with dial plan on gateways

you have things that overlap and it's

not apparent and obvious why they don't

work so you've got to pick them apart

like this this is actually a really good

example so let's run that test again

nine five five five one thousand so

we're turning it into nine three zero

zero zero zero okay that's interesting

we matched rule three again an overlap

problem so let's get rid of rule three

no rule three

once we finally hit that correct rule

I think we're gonna like our results

much better much better liking the

results so our original number is nine

five five five one thousand and our

translated number is nine six one four

five five five one thousand

um I forget if I put a one on that phone

I may want to add one more digit in

there let's uh

Show run and let's take a look at what I

put on that iPhone DN

oh you phone dn2 yeah I put a one in

there so actually

I'm going to um

I'm gonna change the digits here

and we're actually going to put the one

in there so show run we're gonna pull

that translation rule stuff up here so

there's rule four

we're gonna modify that here we're gonna

put an additional one here

that should do it

voice

translation rule one and then rule four

I wonder if it'll let me update it

whoops too many

too many too many too many let me do

this paste there we go

it did let me update it let's run that

test again

there we go now we're giving the nine

six one four five five five one thousand

so I'm gonna make the test call on my

PRI here so nine five five five one

thousand nine five five five five one oh

oh and make the call and there we go

we've expanded it to the ethon dn2 and

we're ringing the phone with nine one

six one four five five five one thousand

so very very cool

um

I think we've shown you the power of

these voice translation profiles and

translation rules and I want you to

experiment with them on your own and

create some translation scenarios that

you would use to do some goofy things

you know you're going to run into these

kinds of problems more often than not in

a production environment and it's

important to understand

exactly what's going on so I think with

that we've covered the critical

components of implementation of dialing

and you know we talked before about how

diopheers match in other videos so I

think you've got a good understanding of

how that's functioning

um

we've talked about you know outbound

preferencing and and how to do those

kinds of things

um I think all of our inbound matching

we've covered we've talked about

examples for

prefixing and stripping and doing regex

really the only thing I think we haven't

covered yet and I'm going to save it for

the last video in this sequence because

it's a topic of its own

um and in fact I've got some really good

coverage of this in my CCNA voice class

and that is class of restriction and

really you're more often going to use

class restriction or Cor in a CME type

environment not so much in the role of a

voice Gateway but I'm not going to go

any deeper than that we're going to save

that for the last video

so for now play with regex play with

voice translation rules and profiles and

inbound digit stripping and outbound

stripping and prefixing and forward

digits and you'll have a really good

understanding of the Core Concepts

necessary to successfully program

gateways manipulate digits and take you

through the CCNA voice as well as this

course the C voice exam so thanks for

watching and I'll see you in the next

video

[Music]

thank you

[Music]

in this module we're going to follow up

with the last piece of dial plan

implementation logic that I want you to

understand and that is implementing

class of restriction on a Gateway or CME

so what I want you to focus on in this

video is the CME portion of things and

in fact what I'm going to show you is a

snippet a video that I did as part of my

CCNA voice video series I think it was a

really good example and fits this

lecture well

so there's no sense Reinventing the

wheel we're going to just go ahead and

launch into that content and talk about

dial plan implementation and class of

restriction on CME

all right so what I want you to keep in

mind when understanding core lists is

the concept of an incoming in an

outgoing core list

now there's more than one way to

configure this and I'm going to show you

one common example keep in mind that

there are

you know you can do it backwards of what

I'm doing you could use core lists if

you wanted to to restrict the inbound

calls to certain destinations but the

example I'm going to show you you know

pretty straightforward

going to control who a directory number

can call

all right so check this out the first

thing we're going to do is go to

convictee

we're going to say dial here core custom

we're going to create a tag called 911.

so name911

we're going to create a tag name local

and we're going to create a tag name LD

incoming core lists

or core lists applied in my example keep

this in mind in my example

are going to be applied

one or more tags

the tags applied are going to be matched

against the tags required by an outmount

core list so let's go ahead and create

the core list

and apply some tags so exit we'll say

die up here core list and we'll create

the incoming ones first and actually

let's create the outgoing ones first

keep it simple we're going to call it

outgoing 911. this is going to be

applied to the 911 so we're going to say

member 911.

exit we'll say dial up here core list

outgoing local

and we'll say member local

and we'll say dial up here

just type it right here core list

outgoing

LD

no space there we go member

LD

so we've created

three

core lists and we'll show them to you

here

outgoing 9-1-1 outgoing local and

outgoing lb these outgoing core lists

are going to be applied

to the dial pairs so for my dial Piers

config T dial peer voice 911

pattern we'll say core list outgoing

outgoing 9-1-1

and we'll do the same one for 912 here

so those are our 911 core lists

for dial peer voice 101 pots which is

our local calls our seven digit calls

we'll say core list outgoing

outgoing local

for 102 and 103 which I'm considering

long distance we're going to assign the

outgoing LD

all right perfect show again show run

pipe again dial up here

can't see what I'm typing here

um dial

peer core

there we go so

we've got a dial pales programmed now

we've got our outgoing core lists

programmed now remember the outgoing

core lists in this example are being

applied to the dial piers

we're going to go and create our

incoming Court lists

50 and we're going to say dial peer

core list

incoming

local

I'm going to say member actually you

know what let's do 911 first just to

keep it in order

incoming 9-1-1 and we're going to say

member

911

I'm going to say dial up here core list

and coming

local and then we're going to say member

911 member

local so you can see that this is

additive so I'm on the directory number

which is where we're going to assign

these going to tag is 911 or local or LD

for you know depending which chord list

I'm in I'm going to say dial up here

core list and command LD we're going to

say member 911 member local and member

LD

so let's go to our iPhone DNS exit we'll

say ePhone dn1 we're going to allow this

guy to dial 9-1-1 so we're going to say

core list and coming

incoming 9-1-1

iPhone dn2 we're going to allow him to

make local calls

iPhone dn3

we're going to allow to make long

distance calls

and that pretty well sums up what you're

going to need to understand from a core

list configuration perspective for the C

voice exam I stopped shy of walking

through any kind of interactive testing

and such because like I made notes on

the screen those diopiers weren't

optimal you know we would have had some

overlaps and problems with call routing

there but uh you know this was more

about core lists not so much to dial

pairs so if you can walk away from this

and understand that core lists are used

by CME to

um create some call routing restrictions

and enforce those on an E phone DN level

you should be good to go for the C voice

exam so with that I'm going to say

thanks for watching hopefully this

information was informative to you we're

going to move into the next video really

into the next section even and start

talking about Gatekeepers and then Cisco

unified border element or cube which is

a session border controller so lots of

fun stuff to come thanks for uh

continuing to stay tuned I know some of

this uh you know it's exciting and

awesome other you know it's a little

Theory dry but we'll do the best we can

to keep things upbeat here so thanks for

watching good luck with your studying

and I'll see you in the next video

[Music]

thank you

in this module we're going to be

discussing a feature set we call Cube

Cisco unified border element cube is one

of those things that's

um to many people anyway seems fairly

new however the base functionality of a

cube and what a cube does has actually

been around for quite some time we're

going to jump right into things and I'm

going to give you a basic Cube primer in

this video and following this up we'll

get into some configuration exercises

show you exactly how to implement and

leverage cube in your environment

like I mentioned before Cisco unified

boiler element is the name of the

product and you may know Cube by its

feature set description and what it does

or maybe even what Cisco used to call it

which was the IP to IP Gateway in fact

if you go back to like iOS 12 and you

know five six seven years ago you know

I'm guessing off the top of my heads

here the first time I heard this

technology mentioned we were calling it

IP to IP Gateway but these days we call

it cubes let's go unified border element

and what cube is is it's a session

border controller and cube is a gateway

I don't want you to think cube is a

separate product you go buy although

you know everybody knows that Cisco is a

licensed happy company and these

features are things you have to purchase

licenses and entitlements for but Cube

itself runs on a traditional voice

Gateway

and differs primarily in the uh the way

that in fact I'm going to show it to you

that both dial piers are VoIP

so think of cube as a feature set and

not a product that doesn't mean it's

free but cube is feature enablement of a

iOS Gateway that you already have

what what are Cube's main

responsibilities well first and foremost

Cube as a session border controller

terminates RTP and rtcp so it's

terminating sip in h.323 IP data streams

it provides Border interconnection

Services and it's a session border

controller so what's this mean to me

well one very popular scenario for where

you would deploy a cube is at your

perimeter if you're doing business to

business internet working between your

IP network your IP enabled Voice network

and someone else's IP enabled Voice

network you'll typically interconnect

these systems via a cube at your

perimeter or perhaps one at both

perimeters

when we look at Cube functionality there

are a number of different core features

and pieces of functionality that Cube

are providing and this is just a short

list there are other features that Cube

can provide you but first and foremost

like I mentioned h323 and sip signaling

Internet working so this lets me

have my call manager

signal my Cube and your co-manager or

perhaps even your Cube signal my Cube

and it's a demarcation barrier so it's

almost like a voice firewall

it provides for DTMF Internet working

you know because we're looking at

different kinds of VoIP dial piers

one Diop here may perhaps be RTP nte or

RFC you know based and the other may be

h.245

alphanumeric and we need the ability to

convert between the two

so DTMF internetworking

codec transcoding just like you know

another Gateway is providing codec

translation and transcoding services you

know Cube can continue to do that

address and Port translations for

privacy and address hiding this is a

security feature but also an Internet

working feature maybe we've both got 10.

networks and they overlap but we need to

be able to internet work with each other

so long as our Cube or cubes have public

facing IP addresses we can signal to

cube and it can signal to us and it can

handle dealing with those Network

discrepancies

and Cube can also provide called detail

reporting normalization for billing

features on a VoIP enabled Network

so when we really start looking at Cube

versus the traditional voice Gateway

what you'll see with a traditional voice

Gateway is that it interconnects an IP

network with a non-ip network such as in

this example a pstn

our Gateway would typically have a VoIP

Diop here

commonly h323 or sip but perhaps it's an

mgcp controlled device as well

and it would have pot style peers

connecting to analog or Digital

telephony Services so that's your

traditional voice Gateway

Cube doesn't look a whole lot different

except that it's an IEP Network on both

sides

so we've got either an h323 or zip

connection to One Network

and an h323 or sip connection to another

Network and a cube is going to have

voice over ipdial peers on both sides or

for both call legs

Cube capabilities you know can coexist

with normal Gateway functions you know

this is just enabling a gateway

to do something it wouldn't normally do

and that is terminate an IP colleague to

another IP colag

when we look at the role of cube in the

Enterprise it can be used to service

internal connections and the reason we

would do this you know perhaps I have

two call manager clusters and I'm going

to interconnect them with a SIP trunk

but I'm actually going to run the SIP

trunk to a cube so call manager a has a

SIP trunk to Cube and call manager B has

a SIP trunk to cue I can now Leverage

The Cube as a point of demarcation I can

do things like digit manipulation Etc

externally we can use Cube as a point of

service demarcation and it gives us

capabilities of address hiding and ease

of integration

with external business to business

connections

more about Cuban the Enterprise there

are a number of different features and

I'm going to show you over the next two

slides many of those features but there

are some that I have omitted from the

list to keep it a little bit on the

brief side you know two pages was enough

but protocol support like I mentioned

previously both h.323 and sip are

supported protocols for a cube the cube

is capable of doing Network and address

hiding which is both an interoperability

convenience as well as a security

feature so we can handle hiding IP

networks and Performing these address

translation functions

Cube does support call admission control

leveraging RSVP which let's be real

hardly anybody uses you know this is

pretty much a non-rsvp world but it is

there and it is supported if you're in

an environment that requires it and you

can use this type of call admission

control to manage the maximum number of

calls per trunk

you know among other RSVP stuff that you

know just normal RSVP

we've got protocol and Signal Internet

working capabilities for IP protocols to

IP protocols so h.323 to h.323

h.323 to sip or sip to sip

we've got media support for RTP and rtcp

and we have two media modes we can

leverage with Cube and I'll show you a

drawing of this in another slide coming

up here but we call them media flow

around and media flow through

Cube can support video with h.261 h.263

and h.264 and it supports transport

modes of TCP or UDP as well as TCP to

UDP internet working

additional features and functionality

provided by a cube DTMF Internet working

so maybe like I mentioned before

one of my VoIP dial pierces using h245

alphanumeric and one of them is using

RFC 2833 we can support that so DTMF

internetworking we've got support for

fax and modem relay and pass-through and

various combinations we've got

supplementary services for hold and

transferring forward using h450 we've

got hold and transfer sub services for

Sip and then we've got the h323 to sip

supplementary service support for things

like media termination points

um Cube can provide for Nat traversal

and do stateful not traversal it can do

qos marking and remarking for IP

precedence and dscp

it has multiple codec support and can

transcode between codecs

uh it gives you security capabilities

with ipsec and secure RTP and TLS and

there's you know other features that I

haven't got into but this kind of you

know hits the high point of talking

about some of the capabilities of a cube

so it's more than just a Gateway it's

kind of like a Gateway on steroids

when we talk about Cube protocol

internetworking we've got support for a

couple of different modes and there are

different features that are supported

depending on which mode you're using so

for h323 to h323 Internet working so you

know I've got two networks that are both

h323 and are connecting with Cube

I can support all combinations of h323

fast start and slow start

if we're doing h323 to sip I can support

h323 fast start to someone doing sip

early offer and I can support h323 slow

start and sip delayed offer in

combinations

for sip to h323 I can do sip early offer

to h323 fast start sip early offer to

h323 slow start and sip delayed offer to

h323 slow start and we talked about

these techniques at length in a previous

video in this course so these should be

you know still something in the front of

your front of your head

and with SIP to sip we support all

combinations of early and delayed offer

so lots of capabilities here for

protocol internetworking

pretty cool

um

thing called media flow around or media

flow through and really what you're

going to choose to use is kind of

depending on

why you're using a cube

with media flow around

let's let's take a look at an example

Network here I've got a cube

with two call manager clusters that are

connected

and let's presume that these are both on

my network

or they're on a common infrastructure

that can communicate end to end

well with media flow around the cube is

not involved in the RTP portion of the

call so the call manager will signal to

Cube for call setup and Cube will signal

to the other call manager

but in the end the IP phones will

exchange media directly that's what we

call Media flow around it flows around

the cube

and it's pretty lightweight on the cube

it doesn't tax a lot of resources

media flow through in a similar Network

topology again signaling between the

call managers and Cubes but the audio

stream from the iPhone is also going to

flow through the cube hence media flow

through this does present a higher CPU

load on Cube than just doing signaling

but again if you're doing

you know address hiding and this is

sitting at the perimeter of your network

this can be a very useful feature to

have because there's a good chance that

the endpoints at either end can't talk

to each other except through the cube so

hence RTP streams flowing through the

cube so very very cool

Cube can either support codec

transparency or codec filtering and it

really depends on what you're trying to

achieve but basically you know keep in

mind cube is just a router with the

capability of interconnecting you know

multiple VoIP networks you know via the

dial peers so the same kind of codec

selection choices that you have

available traditionally continue to be

available with Cube

if you use transparency the endpoints

are going to directly negotiate their

codecs but if you use filtering you're

capable of negotiating a preferred codec

so if I look at ipnetwork A and B they

both support g711 Eula and g729b but the

cube is configured in such a way that it

is only going to internet work with g711

Yule law so we're going to force the

calls to use that particular codec now

this can come in handy when you're

dealing with

an internal network with lots of codec

support

and you're using Cube as a demarcation

point for a carrier provided SIP trunk

perhaps to the pstn and maybe this

carrier only supports one codec so we

can enable you know some negotiation of

these codecs

and finally Cube supports rsvp-based

call admission control again not a lot

of people use an RSVP out there I

wouldn't call it dead but I am saying

that it's not as popular as some of the

other methods

but Cube can support both audio and

video calls and leverage RSVP

capabilities for cat

if you're doing RSVP based CAC it does

require at least two RSVP peers now that

should be pretty obvious to anyone who

is supporting RSVP in your network

because obviously resource reservation

is is kind of a collaborative you know

interactive process if you don't have

multiple endpoints to talk it it's not

going to dig any good so it does require

at least two RSVP peers and when you're

doing RSVP based CAC media flow through

is required

so this is high points on what cube is

and what it can do for you in the next

video we're going to get into describing

some of the cube call flows and exactly

how things work and then we're going to

jump into some setup and actually do

some basic Cube provisioning so I'm

going to leave it at that and I'll see

you in the next video and we'll play

around with the Cisco unified border

element in the CLI so good study thanks

for watching and I'll talk to you soon

thank you

[Music]

you'll have to forgive me this evening

it seems like whenever I'm super excited

about an episode or a video to record

something pops up either I get sick I

get the flu I get the hiccups or

whatever

tonight is one of those nights I've got

the hiccups so I'll apologize in advance

I'll try to try to keep from being

annoying and skipping a beat there

during the video presentation but

tonight I want to talk about Cube and I

really don't want to wait I I want to

kind of go through this

we've gone through Cube fundamentals and

we have covered the what is Cisco

unified border element and we've covered

what are the features and what kind of

things can you do with Cube I want to

walk you through a basic Cube

configuration and that's really what

this video was all about

I want you to think about Cube call

flows and I want you to consider that we

have options for media flow through and

media flow around and you'll want to

select the configuration most

appropriate to your environment you know

based on what your goals are for cube

I also want you to consider the h323 to

323 the 323 to sip or the Sip to sip

feature support necessary for your

environment if you'll recall

um you know relative to support for

things like early offer and early media

Etc

you know there are different rules that

apply in different things that are

supported and that are not supported so

you know keep those in mind and uh you

should be good to go you know keep in

mind three three to three two three

any combination of fast and slow start

on both call legs is supported sip to

sip and in all combinations of early and

delayed offer are supported on both

colleagues

three two three to sip fast start to sip

early offer and h323 slow start to sip

delayed offer is supported and sip to

three two three sip early offered a

three two three fast start separately

offered a three to three slow start and

sip delayed offer to three degrees low

start are supported

when configuring Cube you want to keep

in mind this is just like any other

voice Gateway we've got an inbound dial

up here we've got an outbound dial up

here the same rules we leverage to match

a Diop here applies

and uh you know really the only thing

that's unique is that both call legs are

IP based

in the example we're going to show

here's the architecture of the lab

environment or the the nodes that we're

using in the lab environment anyway

we've got the cube here in the middle

the big blue box it's a iOS router in

this case it's a 2811 it's that HQ

router we've been using and we've got

two IP networks and really I'm showing

you a capability of cube that you know

you may use it for you know this is an

Enterprise kind of model but the same

model could apply equally to an external

interface of cube so I've got two call

manager clusters and I want to build a

SIP trunk between them

but I want to run it through Cube to

give me an additional point of

demarcation perhaps they live on the

same network and then they're in the

same data center but they're managed by

two different business entities or two

different you know arms of a business

that are under separate management

domain

so we've got 10 fan 2 10 10 which is a

call manager cluster and 1010 to 1080.

I'm going to go ahead and pull down a

terminal window here

and we'll make this a little bit bigger

and we're going to walk through basic

Cube configuration now this is that same

2811 we've been doing a lot of testing

on and I've ripped out pretty much all

of the voice configuration you know CME

is gone the um dial piers are gone you

know it's pretty much a router at this

point

the first thing you're going to do when

configuring cube is configure protocol

internet working

to configure protocol Internet working

you can turn on whatever features you're

trying to use and it all starts with

voice service VoIP

now I want to show you all of the

different options available to you under

voice service VoIP and I want to draw

your attention to the second one allow

connections allow call connection types

these this is really where we're turning

on Cube features I'm going to say allow

connections and I've got a source or a

from type protocol selection so I'm

going to say allow connections

sip to sip

and I'm going to say allow connections

sip to h323 and allow connections sip

actually h3232 Sip and allow connections

h323 to h323 so I've turned on

you know kind of the whole Matrix of

protocol support and you'll see it here

under voice service VoIP

that we've turned on those protocol

internetworking components

and the next thing we're going to

configure

is DTMF relay and DTMF relay is one of

those things that's going to happen on

the dial period really you know we're

going to configure more than DTMF relay

we're going to create the dial pairs you

know I could I could assume they're

already there but we're going to go

ahead and create them from scratch here

let's go config T we'll say dial pure

voice

100 VoIP and we'll say destination

pattern you know five dot dot dot it's

an example and we'll say session

protocol

session

Pro session Proto hang on what am I

doing wrong here sesh I gotta spell it

right session protocol sip V2 session

Target ipv4 colon 10 10 210 10. so

that's the first call manager whoops

what have I done wrong here session

Target

ipv4 colon 10 10 2 10 10.

it doesn't like something here what have

I done that it doesn't like

hmm session

session I keep missing the I key on the

keyboard there Target

ipv4 should be fine 10 10 2 10 10. well

that time it worked yeah maybe I had an

extra space or something in there anyway

we've set our session Target

and we'll do a show run here and I'll

get to the bottom and I'll show you the

die up here we've created this is one of

two VoIP dial pairs that we're going to

be using to point to our common inter

clusters and you'll see there it is dial

up your voice 100 void destination

pattern five dot dot dot session

protocol sympathy 2 and session Target

ipv4 1010 210 10. now let me go ahead

and create another one config tile up

here voice 200 VoIP

destination

destination pattern and we'll say six

dot dot dot session protocol

zip V2 session Target ipv4 colon 10 10

210.80 so that's the other one

so we've got two dial piers and this is

pretty much a vanilla uh you know kind

of configuration on the Gateway now if I

want to configure DTMF methods

and let's say that

um

one

of my

connections well you know they're sip to

step so I'm going to back up for a

second because they're both sip in this

example I'm probably going to be using

the same DTMF relay method with SIP

you've got two choices you've got RTP

nte which is RFC 2833 base DTMF relay

and that's an in-band etmf relay method

and with SIP you've also got sip notify

which is

um an atoman method

let's go ahead and say on this one that

we want DTMF relay RTP and te

and then we'll go back to that other

dial up here 100

and we'll say DTMF relay

I'll say sip notify so there you go

we've got a cube we've got our dial

pairs configured

voice service VoIP will allow sip to sip

which is the example that we're doing

here I'll get down here and show you our

dial peers there's our dial peers one

dial up here or one call manager

clusters using sip notify the other

one's using RTP and te

enjoy your Cube It's really that simple

I mean cube is

a gateway

there's nothing all that crazy different

in building a cube to building a Gateway

now there's other things we haven't

configured here we haven't configured

our media type so let's go ahead and

let's let's tell this one for whatever

reason then it's going to be a media

flow through

so let's go to

let me get in here real quick

it's going to be on the dial up here so

config T and say dial pure voice 100

VoIP and we'll say media flow through

and then we'll go to the other dial up

here

and media flow through well there you go

now you got a cube that supports media

flow through I mean are you getting the

picture how easy this is

um

if we wanted to configure

h323 it really wouldn't be all that

different in fact let me do Show run

pipe again dial here voice we'll show

you the existing dial pairs here's 100

and 200. we'll go ahead and convert 200

let's let's say that for whatever reason

200 is now a 323 Gateway so we'll say

dial pure voice 200 VoIP

VoIP and we'll say session

protocol

actually I'm just going to say no

session protocol

sip V2

and that'll peel it back to standard

h323 we can leave the zip session Target

the same

um since we're hp23 we're going to need

a DTMF relay type that is supported

under h323 and you'll notice I've got

different options here because I'm 323

and because I understand what Cube

supports I'm going to say h.245

alphanumeric

so now you've got a cube

Show run

that's supporting

um h323 or 323 to h32 or three two sip

um you know either way depends who's

originating the call one of them is

using sip notify and the other is using

h.245 alphanumeric and Q will take care

of the rest

so what are commands you can use to

troubleshoot Cube well one of the

commands show

and there's no calls active but we'll

see what it would be show call active

voice

you know same kind of command we're

going to use on any other kind of

Gateway

show call history voice

another good one you know there's not a

whole lot going on on this Gateway

so dial pure voice and I like throwing

summary on there at the tail end to see

what dial piers are in there but let's

say I didn't put summary on there you

know we've got a lot of

detailed information relative to the

individual diopia configuration probably

more than you thought you'd ever need to

know if we have calls in progress a show

VoIP RTP connections is useful and

there's no active connections on This

Server right now you've also got some

debugs you can run debug

VoIP IP IP Gateway you know can give you

some good Cube debugs

um debug

cch323 all is useful debug CC sip

messages that's kind of my new favorite

debug of all type good for debugging sip

if you need to get into the h.225 and

h.245 there are various debugs for that

so debug

h225-225 question mark you've got

Advanced q931 ANSI Etc or h.245 you've

got asn1 events or srtp so we're covered

there

um and then really the other big one

debug voice the one I hate more than any

other debug in the world CC API and out

um will tell you everything you never

wanted to know about what's happening in

your Cube or in your Gateway in general

so that kind of cover

level of basic Cube figuration I told

you it wasn't

and it's not all that different from a

normal Gateway you've just got to think

about it a little bit differently you

know certainly there's lots of features

you can leverage that we haven't touched

on they're well outside of the scope of

the C voice exam but they are things you

should spend some time getting your arms

around so I would encourage you to spin

up a couple of call manager clusters

build an h323 Gateway and a SIP trunk

and set up a cube and try to Route some

calls between these things and play with

the different DTMF relay options and the

media flow through versus flow around

and you know if you really want to get

into some interesting stuff when you're

integrating with the cube to for example

an itsp or an IP telephony Solutions

provider or service provider you know

think of it as a Telco over a SIP trunk

you know you're going to find that they

have all kinds of different rules of

what they'll let you do on the SIP trunk

relative to Cube and you'll get into

things like modifying sip diversion

headers and you know for call transfers

to work and all kinds of lovely stuff so

um a lot of fun to be had with Cube lots

of features but yet not really that

complicated to set up so hopefully this

video has been informative to you and

has given you a good summary of Cisco

unified border element and what you need

to understand relative to the C voice

exam in the next set of videos we're

going to talk about Gatekeepers

um the technology that seems like it's

been around for a long long time but yet

hardly anybody is still using because

there are better and smarter ways of

doing things but I digress I'm not gonna

pass too much judgment on Cisco I can't

uh can't be a hater for having options

with technology so anyway thanks for

watching good studying and I'll see you

in the next video

[Music]

thank you

[Music]

foreign

we're going to give you an introduction

to Gatekeepers now I've got to admit I

hate Gatekeepers I I really do

um Gatekeepers come from a Time

when h323 Ruled the Land and we used

Gatekeepers

as a way to globally distribute

h.323 dial plans and endpoint

information think of it as kind of like

a dynamic routing protocol where we're

routing by rumor and saying I've got

these guys and another router says well

I've got these guys

it worked

it's still used somewhat but from a

design perspective we tend to do things

a little bit differently so you're not

going to see as many Gatekeepers in

modern networks that you would have say

15 years ago or 10 years ago even so

times have changed a little bit but that

said understanding Gatekeepers is still

part of the C voice exam and it's

something you're going to need to

understand particularly in larger

Enterprise networks so let's Jump Right

In and talk about gatekeeper technology

how it's used in a little bit of the

fundamental concepts and then in the

next video we'll go through some basic

gatekeeper configuration

first off understanding the role of a

gatekeeper what a gatekeeper does is it

allows us to globally distribute an

h.323 based dial plan so for example

we're going to be able to advertise one

endpoints we have registered to us and

it allows for local dial plan control

and what I mean by local dial plan

control is that a a given Gateway

can have its resources or not even a

Gateway a given

um call manager for example can register

with a gatekeeper you know a given

endpoint

or you know group of endpoints can be

shared

with you know from a gatekeeper to

another GateKeeper so I'm kind of

rambling here but here's a good drawing

it demonstrates it way better than I can

explain it so let's say I've got two

um you know two gateways here and a

Gateway is you know let's say these are

CMEs or whatever and we've got endpoints

on them you know the left one has you

know five thousand one five thousand two

and five thousand three extensions and

the one on the right has you know one

thousand one one thousand two one

thousand three and up here in the middle

we've got a gatekeeper so each Gateway

can tell the gatekeeper what endpoints

it's in control of it and you make the

gatekeeper aware of where endpoints are

on the network

what are the responsibilities of a

gatekeeper well there are both mandatory

and optional responsibilities of a

gatekeeper the mandatory

responsibilities on the left are going

to include things like address

resolution so if I'm an h323 Gateway you

know and I've got a call coming in from

the outside

and

um you know it may have an e164 address

or an alias assigned to it that we're

going to use to route that call so

address resolution you know of those

addresses to an endpoint would be one

role of a gatekeeper admission control

bandwidth control Zone management those

are all mandatory responsibilities of a

gatekeeper and those are the core things

that they're going to do optionally a

gatekeeper can do call authorization

based on policy you know maybe some time

of day policies it handles call

management so if an endpoint is busy it

can redirect a call perhaps and it can

reject calls you know do bandwidth

management if insufficient bandwidth

exists so lots and lots of things can

happen within the realm of a gatekeeper

and like I mentioned here a call manager

cluster can register with a gatekeeper

endpoints non-call manager can register

with a gatekeeper we saw a lot of video

endpoints doing this back in the day and

each endpoint can be registered in one

zone and we'll talk about zones when we

get into actual configuration of

Gatekeepers when I'm going to kind of

keep that out of this video for now

when we talk about the gatekeeper call

Process endpoints are going to register

with the gatekeeper at startup so you're

either going to go through an auto

Discovery process where you're going to

be programmed with the IP address of

your gatekeeper and we'll talk about

that you know those options here in

another couple of slides but you're

going to register with the gatekeeper at

startup and when you want to place a

call you're going to request admission

you're going to send a message to the

gatekeeper it says I want to place a

call and it's going to look and see can

you place a call to the endpoint you're

requesting to place a call to is it busy

is it available do I have any idea where

it is and it's going to send you a

message back and either you know accept

that request and confirm it or reject it

and more about that as we go

if the gatekeeper accepts the request

it's going to return to you or return to

the originating endpoint a destination

IP address so you're going to know where

to go next

gatekeeper signaling

I can tell you right now this is

something you're going to have to

memorize

um this is at least you know the tests

I've taken your knowledge of this is

something that uh they expect to be very

thorough so hint hint good test material

here Gatekeepers use h.225 Raz for

signal laner for the initial setup

that's going to be over UDP

next h.225 over TCP is going to be used

to set up the connections between the

endpoints and if there's no gatekeeper

present that htt5 signaling is going to

just happen between the endpoints

directly once we've established a

connection we're going to use h.245 with

TCP to establish the media flow and then

throughout the duration of the call

h.245 signaling is going to continue to

manage that call

and here are those messages you're going

to need to understand boom I'm going to

walk you through these and give you a

basic overview of what these are used

for so it all starts off with Gateway

Discovery messages

endpoints are going to either use

unicast or multicast to try to discover

their gatekeeper and we've got three

different messages at play here

gatekeeper requests gatekeeper

confirmations and gatekeeper rejects a

gatekeeper request message is you know

plain and simple it's a message that's

going to be sent to the gatekeeper from

the endpoint

you know asking you know it's a

discovery message you know are you there

um a confirmation a GCF the gatekeeper

confirmation is going to be a reply from

the gatekeeper indicating

um the transport address of the

gatekeeper rash Channel

the gatekeeper reject is going to be a

reply again from the gatekeeper

rejecting the request for registration

so there could be you know a number of

different reasons I'm not getting into

for that now but I want you to see a

pattern we've got requests confirmations

and rejects and that pattern is going to

apply to many of these different message

types so as we move on Gateway and

terminal registration we've got

registration requests registration

confirmations and registration rejects

we've got gateland terminal

unregistration you know confirm you know

request confirm reject we've got

resource a availability indicators our

AIS

and then confirmations we've got

bandwidth requests confirms and rejects

location requests confirms and rejects

and the list keeps going yes there's

more we've got and these are really

important ones right here the first

three the admission requests

confirmation and rejects those are what

we're going to use when we try to

request admission to place a call we've

got disengage requests confirmations and

rejections and we've got a request and

progress message and we've got some

information request messages for status

so information request and response and

acknowledge we've got a negative

acknowledgment or a knack and then we've

got an information request confirmation

so lots of h.225 Raz messages out there

again you know it's just stuff you're

going to have to come into memory

um you know and you'll want to read

about all of these different Ras message

types there's way too much information

here for me to get into in a video you

know this is something that you know

merits cracking open a textbook and

going through so

I will not try to explain them all in

granular detail but I do want you to

understand the call and Mission request

the arq's ACs arjs because we're going

to use the heck out of those things

talking more about the gatekeeper

Discovery process I told you before that

endpoints are going to attempt to

register with a gatekeeper when they

power up and what they're trying to do

is you know discover a gatekeeper so

they can register with it and ultimately

discover what zone they're a member of

and I told you I'll talk about zones

when we go through the configuration

video so we're going to leave it at that

level of detail for now this discovery

process can be either unicast or

multicast and this is one of those

message types the Gateway I'm sorry

gatekeeper request message the grq that

we're gonna send

if it's unicast or if we're doing a

unicast discovery we're going to use UDP

Port 1718 and it's going to require

static configuration of the gatekeeper

on the endpoint so I mentioned using

this on video endpoints years ago we

would you know literally we'd program an

IP address a subnet mask a default

gateway and another one of those

parameters we would set up is gatekeeper

IP address so that was a unicast based

discovery you can also do a multicast

based discovery which is going to use

multicast address

224.0.1.41 and if you're doing

multicasting for gatekeeper Discovery it

is not necessary to have static

configuration and your endpoint it's

just going to use multicast

and you know when you send this

gatekeeper request message you're either

going to get a GCF which is obviously

what you hope for a confirmation of

gatekeeper um registration or a grj for

a gatekeeper rejection

and again that can be a number of

different reasons usually these are

configuration related errors but uh you

know other things can happen there

gatekeeper re-registration back early on

with h323 prior to version two so we're

talking h323 version one a gatekeeper

and an end point there would be a

registration event take place every 30

seconds you know in full effect you know

um

it would basically be a gatekeeper

request gatekeeper confirmed gatekeeper

request gatekeeper confirm every 30

seconds

well that was kind of overhead on a

network that really wasn't necessary so

in h323 version 2 we came out with

lightweight registration and what's

going to happen with a lightweight

registration is you're going to send a

registration request just like you

normally would and it's going to contain

a TTL value if it doesn't

um you know the RCF coming back to you

is going to have one added into it and

it's going to return you a TTL value

shortly before the expiration of this

TTL value the endpoint is going to send

an updated rrq message with a keep alive

field in that message set to true so

we're shortening the process here and

you know obviously if you know you

expire those ttls you do it you know

eventually the thing is going to fall

off the map but uh

you know more about that later

let's talk about a couple of these

messages in detail the more critical

messages the things that are going to

make or break base gatekeeper

functionality the first one is the call

admission stuff the arq which is an

admission request an admission

confirmation ACF and an admission

rejection arj so to place a call the

endpoint is going to send an admission

request message to the GateKeeper the

gatekeeper is going to look at the

endpoint check its status see if it's

busier available or you know able to

have a call routed to it and the

gatekeeper is going to return a

confirmation or a reject message so

you're going to send an arq and you're

going to get back in ACF or an arj

un that simple um you know that's your

call setup process

information request messages this is the

last one I want you to have an

understanding of

um they're sent periodically to all

registered endpoints this is basically

you know endpoint death detection you

know it's sent to verify that the

endpoint still exists

um if we

don't receive a response and information

request response the irr uh the endpoint

will be considered dead and aged out of

the system

with that we've covered basic gatekeeper

Concepts and or at least the high level

stuff anyway we're going to get into a

little bit more we'll talk about zones

and configuration in the next video and

that'll pretty much wrap up this section

of the Course once we get through that

you know there's not a whole lot left

we're going to talk about quality of

service and you should be in pretty good

shape for final preparation towards your

C voice exam so I'm going to wrap this

one up I'll see you in the next video

we're going to do some configuration to

Gatekeepers and then we'll move on to

qos thanks for watching have a good

studying and I'll see you in the next

video

[Music]

thank you

foreign

we're going to go through some basic

gatekeeper configuration did I mention

that I hate Gatekeepers I wouldn't be

clear it's not because they're difficult

or challenging I just think that we've

reached a point where there are better

ways to do things but if you ever have

the need to configure a gatekeeper to

distribute some h323 dial plan among you

know your Enterprise Network you'll have

the skills and knowledge necessary to do

so now let me start off by saying that

configuring Gatekeepers is really really

easy I mean this is not a crazy process

that is

um you know some big Magical Mystery

it's really pretty straightforward I'm

going to walk you through the most basic

configuration examples and talk a little

bit about how this works

and that should be adequate to cover

what you're going to need to understand

to successfully pass the civilis exam

and really to deal with Gatekeepers if

you ever do cross paths with one

let me go ahead and start by bringing up

a connection to our router here this is

the same 2811 we've been playing with

all course I love this Hardware it works

really well for me and I want to tell

you that I took a little time to install

a new iOS on this router the reason for

that is the code that I was running did

not support the gatekeeper the

gatekeeper feature set and I was a bit

surprised that the IP voice which is

what I typically run on a voice Gateway

that that didn't support the gatekeeper

feature set so I went ahead and went out

to feature Navigator and looked it up

and found an iOS that did support it and

I've gone ahead and I've updated the

router so we're running a little

different code than we were before but

uh just a hint but when you try to type

these commands in you may get an error

that the gatekeeper command does not

exist so you'll need to visit your

friend feature Navigator and find out

what iOS and feature said is necessary

to support this on your lab platform so

same old router nothing crazy going on

here we're going to walk through some

basic gatekeeper configuration first

thing you're going to do config D just

like anything else and we're going to

say gatekeeper when you do this the

first time is you go through the

configuration it may actually prompt you

in fact it may very well prompt me

uh to accept a Eula for a license

agreement because this is a feature that

you have to turn on so you know if that

pops up you'll understand why so

gatekeeper

Gatekeepers have zones and we've kind of

talked about zones before a zone is a

group of devices under a management

domain

you may have one or more local zones and

you will likely have one or more remote

zones I'm going to configure one local

Zone and one remote Zone in this example

and that should give you what you need

to use as a building block to configure

more

so what I'm going to do is I'm going to

go ahead and I'm going to say Zone

local

and I'm going to say Columbus because

that's where I'm at I'm located in

Columbus Ohio so Zone local Columbus and

I'm going to define a domain how to

network.com

and then I'm going to say 10 10 210.1

that's the IP address of this Gateway

that we're going to use

for our

um for our Sim or for our GateKeeper

um I can create multiple local zones if

I have multiple local zones I don't but

if you do certainly feel free to create

those I'm going to go ahead and create

two remote zones I'm going to say Zone

Zone gotta get my fingers on the

keyboard here remote and we'll create

one called New York

how to network.com and it'll be

10.50.210.1

and we'll create Zone remote Chicago

how to network.com and it'll be

10.100.210.1 so we've configured one

local Zone and two remote zones now I'm

going to issue a no shutdown command

and it actually would have asked you for

the EULA here but I've already accepted

it so we're good to go and it tells me

that the feature gatekeeper is activated

there's no error so we've now configured

the most basic settings related to a

gatekeeper now

the whole purpose of Gatekeepers is dial

plan distribution so we're gonna need

some kind of routing intelligence to

build a dial plan

that understands how to reach other

Gatekeepers not necessarily gateways but

other Gatekeepers to look for endpoints

we're going to do that using Zone prefix

and what zone prefix is going to do is

it's an association between a dialed

number and a Gateway really I mean you

know when you really get down to it at

the end of the day

we're going to do that by going again

we're still on the gatekeeper

configuration mode we're going to save

Zone and let me show you what else is

available to you here

um you know a few different commands I'm

going to go ahead and say prefix and

it's going to allow us to define the

e164 address space so Zone prefix and

I'm going to say Columbus and I'm going

to say five dot dot dot so I'm telling

this gatekeeper that any end points or

any e164 numbers matching that pattern

are in the zone of Columbus

I'm going to say Zone prefix New York

six dot dot dot

and I'm going to say Zone prefix

Chicago two dot dot dot and you know

there's other kinds of wild card

patterns and things you can use but I

think that that gives you the basics

here

let's talk about technology prefixes

that's the other thing you're going to

want to understand when dealing with

basic gatekeeper configuration a

technology prefix is an optional h323

feature and you know the certain skinny

of it is it gives you additional

flexibility in making call routing

decisions

um it's used to group gateways together

by type or class so think of it as you

know if I have some gateways that are

used for voice and I have some gateways

that are used for video and I have some

gateways that you know perhaps point to

h320 and clients or you know there's all

kinds of ways to do this but you know

it's a class or a grouping of devices

I can use technology prefixes to pool

these gateways

into a group

um for example you know that we could

use and in fact we'll go ahead and we'll

build it here one pound let's say that

that's the technology prefix used for

voice if I try to route a call

based on a technology prefix of one

pound I'm going to randomly select the

Gateway

in that pool and there you go

if the majority of your calls are going

to use one class or one type of service

you know like voice what I just talked

about with the one pound

you can also Define a default technology

prefix so let's go ahead and configure

that I'm going to say GW type prefix one

pound star and what one pound star means

is Tech prefix of one pound and any

number that's at least

no digits or longer

um

is going to match that that pattern and

then I'm going to say default technology

so we've configured the Gateway prefix

of one pound star and we've we've

actually made it the default technology

prefix now if you wanted to configure an

h323 Gateway

to

register with your gatekeeper you would

go into the gateway and it's actually a

very simple process it's just part of

the h323 configuration

and what you do I'll go ahead and I'll

do it on the same box here you know

typically you'd be doing this on a

different router but for the purposes of

this demo that'll suffice I want to go

ahead and create a loopback interface

and loop

the 10 I don't know if I have a 10 I

might have a zero so we'll just skip it

and go to 10. we'll say IP address 172

16 20.20.

and we'll bring that into service so

loopback10 goodest place of any to use

it we're going to say h323 Gateway

VoIP interface to turn h323 oh I can't

do that I've already got it turned on on

fast e00 okay you know what I'm going to

go to fast d00 and temporarily turn that

off and FAO no h323

three two three Gateway

um VoIP interface

we'll go back to my loopback and do it

here

I don't really want to mess around too

much oops put it here there we go I

don't want to mess around too much with

that ethernet port

um the next thing we're going to do is

an h323 Gateway

VoIP bind Source address

and I'm going to use the address of that

loopback interface so

172.16.20.20 and I'll hit enter

now this is where the rubber meets the

road here I'm going to say h323 Gateway

VoIP

ID

and it's going to be the ASCII ID string

of this Gateway so I'm going to call it

Columbus one

and I'm going to say IP address and I'm

going to type in the address of my

gatekeeper so let's say it's 10 10 or

I'm sorry 10 50 to 10.1

and we'll go ahead and hit enter

and I'll again Define

um my h323 ID and we'll go ahead and

call that Columbus so h323 Gateway

VoIP

um

h323 ID Columbus enter

and if we wanted to define a tech prefix

we would then go VoIP Tech prefix Tech

maybe I can't do that on this iOS I

should be able to do that oh no I'm

missing the command h323 VoIP Tech

prefix one pound

so that's pretty much it I mean that's

how you tell a gateway to register with

a gatekeeper you know if you were

configuring dial piers in fact I'll show

you what a dial up here would look like

just because I'm in here very simple to

set up you know dial pure voice we'll

call it

500.8 destination pattern and we'll say

something like

um you know five dot dot dot

and then I'll say Tech prefix one pound

session Target Razz

session

Target rats enter so now we're going to

use the GateKeeper to Route the call so

that's pretty much it

um you've got a couple of commands you

can use to verify that your Gatekeepers

working I don't have any endpoints

registered to this one so the output

will be a little bit uh on the blank

side but some commands that are useful

for you is show GateKeeper

in fact let me stop there let me show

you all the commands are down here

you've got calls circuits clusters and

points we've got the type prefix stuff

performance data server status Zone all

kinds of stuff so if I go show

gatekeeper GW type Tech or GW type

prefix we'll see the prefix that's been

configured and that it's the default

gateway technology

we can do a show key keeper status

and it's going to show us that the

gatekeeper is up and it's going to give

us other information about the

gatekeeper you know if you've done any

kind of bandwidth stuff with it you know

you'll see those values here a show

keeper Zone prefix

it's going to show us Zone prefix table

information that includes the gatekeeper

name and the e164 prefix and use

if there are calls that a gatekeeper is

aware of you could do and we won't have

any but you could do a show gatekeeper

calls and it would give you information

obviously we have no active calls

if you want to know about the endpoints

registered with you as a gatekeeper show

gatekeeper I'm going to let you guess

here what do you think it is yes

endpoints

um you know I don't have any active

registrations right now

but that's it if you get into debugs

remember we talked about two protocols

that were being used we've got h.225 and

h.245

um you know the normal debug

h225 or h245 you'll see them both listed

there

are useful to you and then you've got

options here for asn1 advanced or q931

you can also debug the Raz messages

which is debugger as so there's not

going to be a whole lot useful here

because we don't have anything going on

but you know that's pretty much it

um

if you wanted to get into call admission

control and deal with bandwidth base CAC

you can do that it's as simple as

defining the bandwidth under the

gatekeeper so I can go config T and

we'll say gatekeeper and then I've got

these commands here for bandwidth and

then you know I could just set you know

bandwidth interzone

[Music]

defaults you know and put a value you

know of bandwidth in there seeing like

256k

and then I could go bandwidth total

default

you know I could say something like uh

you know 10 Meg

so we kind of zeros in there

um bandwidth session default

you know 512 Etc so you can you can do a

lot of things here as far as cat goes

but I don't want to dig too deep

um I want to just give you the basics

and get you started on Gatekeepers and

and provisioning things so remember to

get an iOS that supports it in the

feature pack remember there's nothing

too scary about Gatekeepers and remember

don't put these things in production on

new networks there are better smarter

design methods

um to do this than gateways I've got a

colleague who has 200 000 Cisco IP

phones in his Enterprise and how many

Gatekeepers do you think he has

one

so with that I'm gonna say thanks for

watching

um we've wrapped up this section in the

next set of uh set of modules we're

going to get into ipqos and talk about

qos design strategies and give you some

basics of what you need to understand

for IP quality of service on a unified

Communications Network so thanks for

watching have a good evening and I'll

see you in the next video

[Music]

thank you

in this module we're going to start

diving into qos and IP quality of

service and we really need to start by

talking about what are the quality

issues that exist in the network today

that we need to be considering when

designing implementing and supporting

these unified Communications

environments so let's start at the

beginning and let's talk about where we

came from and the networks of yesterday

and the reason I call them networks of

yesterday is because they they were they

were truly separate unique environments

so we look here at a typical Data

Network where we would have our Lan and

a router or a switch in a Wan connection

and then we'd have you know the same

thing at the other side and these were

point-to-point lengths and this was the

data Network and it worked really well

and then we had the phone system and we

had our pbx's and they would connect via

the pstn and then we had our video

conferencing endpoints you know maybe

they were on an h320 Network and they

were just separate environments voice

and video and data you know were all

using their own equipment and following

their own Pathways and had their own

addressing schemes and you know they

were separate and it worked

but we had all these multiple

infrastructures to manage we had to some

extent redundant types of services they

weren't necessarily redundant Services

because a pstn at this point in time was

not something that was a substitution

for a frame relay connection and you

know it didn't handle

you know you know the video links and

all the stuff you know it was all

separate so you know you couldn't just

swap one out for another and as

technology evolved we began to get this

ability so when you look at the network

of now

it's really not all that different from

the network of yesterday

except for the fact that we've

Consolidated Technologies we've

collapsed things together instead of

having to run video over

h.320 and ISDN lines were packetizing

the content and sending it over our

packet switched network instead of our

circuit switch to network the same thing

is happening with voice traffic and

telephone service we're packetizing this

content and using RTP over UDP and

sending it across the way and you know

through the the networks at the

individual sites our land is you know

our PCS were still using the data

Network

and you know these phones and these

video endpoints Etc they're all just

more endpoints on the land so you know

you get an idea here of what's going on

we're really collapsing things together

and the WAN becomes a facilitator for

all types of traffic to any destination

regardless of mode of communication so

that's the network of today

when we think about voice quality and

delivering a consistent voice experience

over the data Network there are a number

of different factors that we have to

account for

when we look at what voice and video are

there are a number of different problems

that you can anticipate occurring and a

couple of these you know for voice you

know what can you expect your help desk

to hear as a phone call you know you

know I'm placing a call and my audio is

an unintelligible or it breaks up is

choppy and I can't have a good

conversation because I keep losing

pieces from the other end you know

that's one of the conversations you

might hear for someone experiencing

voice quality issues

you know the video stream is jerky and I

keep getting pixels on the screen things

are freezing and you know those are the

kind of calls that you can expect to

hear when video is not being transported

properly across the IP network so what

do we do about these things or you know

what first we got to start my

understanding the fundamental needs of

these types of transport you know of

these Technologies so voice and video

they're time sensitive the the you know

the traffic is only meaningful or the

content is only meaningful if delivered

on time

when you look at packetized voice and

video

we have packets of a constant size

that are competing with bursty data

traffic

and from a user perspective you know

obviously downtime is not acceptable you

know you pick up the phone they expect

dial tone to be there and they expect it

to just work

so these are all things we've got to

consider

when thinking about

quality on an IP network so not only

what are the technical requirements but

what are the user requirements what's

going to be acceptable to the end user

and what's not

some of the technical issues that we

have to resolve

one is bandwidth you must have adequate

bandwidth and you must use the bandwidth

that you have effectively so this

doesn't mean throwing bandwidth at the

problem is a solution but it does mean

that you can only optimize something

that you have enough of to begin with so

if you don't have enough to begin with

optimizing it's not going to do you a

whole lot of good because at the end of

the exercise you're still not going to

have enough so you must be able to use

the bandwidth you have effectively and

you must have enough

we need to be able to manage delay or

latency of data arriving if data arrives

you know too late it's not useful

anymore and we need to manage Jitter

those variations in the delay

we must be able to prioritize the

traffic that is not tolerant and cannot

handle the delay and voice is a great

example of that

and we need to prevent packet loss you

know on our data networks packet loss

was not such a big deal because the

protocols like TCP that we reuse or that

we use

have mechanisms built into them to

acknowledge that data sent was received

and retransmit data as necessary well

that's not terribly useful with a voice

or video type of network because like I

said before this the information is

delay sensitive if it doesn't get there

on time it might as well not have gotten

there at all

so as we go into planning for man with

and this is just you know a summary

we're going to dive into how do we

address each of these issues

specifically in the next slide deck but

as we plan for bandwidth keep in mind

that you only have as much available

bandwidth as the weakest link in the

chain so if I've got a network here with

a a 10 Meg hop and 100 Meg hop and a T1

link and then another 10 Meg keep in

mind that for traffic Crossing this

network

you've only got As Much available

bandwidth for voice conversations or

video conversations or you know whatever

you're talking about

as the weakest link in the chain so if

I've got a T1 line in this otherwise

high-speed LAN

then consider that that T1 line that one

and a half megabits per second

connection

is going to be the weakest link in the

chain and it's going to determine the

amount of Voice or video or whatever

traffic

I can successfully deliver end to end on

my network

so for talking voice you know you take

that make and a half link you divide it

by 80k per stream and you're going to

come up with the number of supported

streams that you can support end to end

on your network based on this weakest

link

traffic classification

is going to be one of the techniques

that we use as part of a overall qos

strategy and again we're going to get

into these strategies and qos policies

in depth in the next video or several

videos

but by classifying traffic and putting

it into buckets

and then prioritizing the delivery of

the traffic appropriately

we can effectively increase this usable

link bandwidth

once we classify traffic properly and

put it into these buckets

we can give it the appropriate treatment

from a queuing perspective and there's

various queuing mechanisms available to

us

we have strict priority queuing custom

queuing weighted Fair queuing

class-based weighted fair queuing and

low latency Q in and we'll get into

these techniques as we continue to go

but I just want to show you you know

some of the technologies that are

available to us

in consideration of how we might resolve

these quality issues on our environment

when we talk about delay and Jitter

which I mentioned before there are a

number of different factors that

contribute to Total path delay and

Jitter on a network you know if I look

at a network here we're going to have

you know things like serialization

delays you know how long does it take to

put the bits on The Wire we're going to

have propagation delays that are you

know just laws of physics stuff you know

how fast can at1 move bits of data from

Los Angeles to New York it takes time

because I'm crossing physical distances

so there's a number of factors that are

going to come into play here

and finally packet loss you know we look

at networks where we're typically

um you know an experiencing packet loss

is going to be at points of congestion

so if I look at this You Know sample

Network here on the screen you know I've

got a high speed Network on one side of

a device you know 10 mega bandwidth and

a significantly lower Speed network for

a connection to the next device you know

one and a half megabits obviously at

that interface where the I'm trying to

shove 10 mega traffic onto a one and a

half Meg interface we're gonna have

packet loss we're going to have to drop

some packets because we can't sustain

that 10 make throughput anymore so we'll

have to think about packet loss within

our Network

so these topics are these you know

bullet points have been very high level

summaries of the types of things you're

going to need to consider from a user

perspective and from a technical

perspective when managing quality on a

voice Network

or a converged Network I should say in

the next video we're going to talk about

qos strategies for unified

Communications networks and we're going

to get into a little more detail on this

traffic classification and then what we

can do from a policy perspective and you

know talk about how and when and where

we should do these things and as we move

forward we will go through some examples

of configuring qos in fact we're going

to have a pretty good section on

actually implementing a qos policy on a

network so

with that I'm going to say thanks for

watching

it's been fun and I'll see you in the

next video good luck with your studying

[Music]

in this module we're going to talk about

qos policies so previously we talked

about many of the qos fundamentals and

how do I you know look at my network and

understand the quality issues that face

me and we talked just a very little bit

about some of the things you're going to

consider when addressing those issues in

this module we're going to dig deeper

into that topic talk specifically about

things you need to Target when resolving

those quality issues and talk about qos

policy design as a whole now I'm going

to go through an example policy in this

video

but it's not the one that I use in

production networks in fact you know

later we're going to cover a video

specifically on The Cisco Baseline qos

model which gives you an 11 class policy

that I've learned to adopt over the

years now I don't necessarily Implement

all 11 classes but I use the structure

so that I have the flexibility as the

network grows and expands but I digress

let's get back on the topic for this

module qos policies now when we talk

about addressing quality issues in your

network we brought up you know delay is

a problem

and talked about the need to reduce the

delay so that's one of the things that

you're going to need to consider when

addressing quality issues so how do I go

about doing this well first of all make

sure you have adequate link capacity or

bandwidth I talked about using qos as a

means to utilize the bandwidth available

to you most efficiently but you have to

have the bandwidth in the beginning to

optimize it plain and simple you're not

going to send two Mega data down a one

megabyte I don't care what kind of qos

you have it's not going to happen so

make sure you've got adequate link

capacity and bandwidth

let's make sure that you're prioritizing

those types of packets that are latency

sensitive so an FTP session is not

latency sensitive a telnet session can

be voice and video most certainly will

be so make sure you're utilizing a

method to prioritize your latency

sensitive packets and we'll talk about

how to do that within this module

I want you to consider the use of

payload and header compression

particularly on low speed links when

you're dealing with fractional T's and

things below you know 768k and I know a

lot of you International guys you know

are going to tell me hey you know in my

neck of the world we don't have

bandwidth like crazy you know we're

using Frac t1s all over the place and if

so that's okay consider these other

alternatives for you with compression of

data

we've talked about the need to reduce

packet loss and we have to think about

really where packet loss comes from

packet loss comes from congestion most

of the time so again back to you know

the same link capacity thing either you

have enough bandwidth to meet your

business needs or you don't if you don't

Pony up and increase your bandwidth so

reduce the congestion points you know if

you have traffic that has a requirement

obviously your network needs to be sized

appropriately

consider increasing your buffering space

and there's lots of buffer tuning you

can do on Cisco routers and this will

help you deal with bursting of traffic

and drop the low priority packets try to

reduce the congestion you know if you

can't necessarily increase the link

capacity use qls methods to drop lower

priority packets you know back to my FTP

example if I'm doing a data transfer

with TCP data and a little bit gets lost

now and again

sure I may slow down the transfer but

I'm not going to kill the transfer TCP

has built-in mechanisms for

re-transmission of lost packets whereas

UDP such as voice does not so if I don't

have enough Elbow Room

selectively drop the data that can deal

with being dropped not the data that

can't

so now that we're gonna you know get

through you know the basics and the high

level topic of what do I do you know

it's kind of leading you back to the

conversation of you need to design an

Enterprise qos policy and the first

thing I want to get in your head is that

when dealing with quality of service you

need to think Global qos is a strategy

that you leverage it is not a feature

you turn on I mean granted it is a

feature you turn on but features don't

solve problems and features don't

Implement qos I don't care how big or

how small you are you need to think

about your traffic and develop a global

Enterprise qos strategy

you need to write your qos policies

considering both the business

requirements as well as the technical

requirements and you know kind of the

point I want to lecture here about a

little bit and I may be on a soapbox a

little bit here but you cannot as a

network architect as a network engineer

you cannot simply let the business

decide what is more important and what

is not you cannot let the business group

things into these qos buckets that we're

going to talk about now that doesn't

mean that their business priorities

don't come into play they absolutely

come into play but you as an engineer

need to translate those business

priorities along with the technical

priorities into a comprehensive qos

strategy so you might as well start now

getting the support of your management

team that the business doesn't get to

just gold silver bronze things there's

more to it than that and you as a

technical resource are going to help

liaise between the technology needs and

the business priorities so you know you

need to be part of influencing these

decisions and I like to say show the

business what they need but listen to

what they think they need

so I know I'll stir up a lot of a lot of

discussion you know based on that

recommendation you know but I've been in

that role you know I worked for you know

major industries where

um you know the business had certain

priorities and they knew what those

priorities were and they were important

to us getting paid and then you know

servicing our customers however

if if I would have simply asked the

business to create a qos strategy and

prioritize data it would have been

completely broken and we probably

wouldn't have been doing business

because the right types of traffic

wouldn't have been getting through so

act as that liaison

when we talk about steps for

implementing qos there's really three

steps in the planning process that

you're going to need to do first off

step one identify the traffic you need

to know what traffic is on your network

and this is not something you can simply

guess at it's not something that you

really should just you know say oh I'm

I'm the network guy I've been here for

15 years I know what's on my network you

know what you probably don't know what's

on your network it changes applications

are doing things behind the scenes that

you don't necessarily become intimately

familiar with for all applications so

take an audit look at your network

figure out what's really going on on

your network and consider both the

business and Technical implications of

what this traffic is

step two group Traffic into classes you

know when you have traffic that has

similar treatment requirements you can

group it together and that's going to

make managing your qos policy a lot

simpler you can't you simply can't

Define 500 kinds of traffic and 500

treatments that's insane that's not

scalable and it's not going to perform

well so you need to create traffic

classes once you've classified traffic

into buckets you're going to define

specific qos policies and definitions of

how you're going to treat that traffic

you know for example I may allocate and

guarantee certain amounts of bandwidth

for a specific type of class I may do

strict priority queuing and you know for

a type of class I'm a

decide that a particular class of

traffic is less important than default

you know it's a scavenger type of

traffic that I only want to cross the

network when nothing else needs to

bandwidth so you'll come up with these

rules and we're going to give you an

example in this slide of you know a very

basic and typical model and then we'll

actually have a slide later on about the

Cisco qos Baseline template and I

remember first seeing that template back

in one of the Cisco press books I want

to say Tim spaghetti was the author and

it was something like Enterprise qos

design or something along those lines

was the title fantastic title Tim's a

great guy knows everything in the world

there is to know about qos you know puts

me to Shame by a long shot but um you

know a lot of those resources you know

have developed these methods you know

we're simply sharing with you at this

point in the game what works you know

and uh what best practices you should

follow so these are the steps you're

going to follow to implementing qos on

your Cisco Network Step One is

identifying the traffic you need to

understand the traffic on your network

and I kind of Dove a little deep into

this you know in the summary but we'll

hit it again it's important that you

understand exactly what's happening on

your network do an audit and don't

assume you know I've found when doing

Network audits in the past that there

was not only IEP traffic we didn't know

about but there were other protocols in

use we had print servers speak in ipx

and apple talk and we didn't have an

apple in the building you know we we

didn't have anything running ipx these

were just defaults these this was

garbage traffic that was on my network

that we went around and we disabled and

we cleaned up and you know reduced some

uh you know some ethernet latency that

we had obviously you know could I

measure it yeah maybe maybe not but was

it there yet there were there were

benefits there so do it on it don't

assume know what's happening on your

network

document the technical and business

requirements for the traffic and

applications on your network if you're

running a latency sensitive telnet you

know transactional based system like a

terminal emulation to a Mainframe

environment obviously that has certain

performance metrics that are going to

need to be met for it to be a usable

application you can't just even though

this telnet data is you know is data

traffic doesn't mean you can slow it

down and kill it to the point that it

interrupts these interactive sessions

however that FTP data transfer or that

backup that is running you know perhaps

you can so understand both the technical

and business requirements for the

traffic and applications that are

running on your network and you need to

also Define the necessary service levels

for each traffic class and I talk about

that a little bit with prioritizing you

know telnet for example you know if

you've got these interactive sessions

they need to be treated with a certain

service level that you know back-end

data doesn't necessarily need to be

treated with so consider both the

business and the technical requirements

step two we're going to group Traffic

into classes and I'm going to say from

my experience somewhere between 5 and 10

classes of traffic is common I mentioned

before the Cisco Baseline qos model with

11 classes I love it you should use it

um but we're going to give you a little

simpler example as we walk through this

I've seen in the past a lot of

Enterprises use this whole gold silver

bronze model for defining traffic types

and in my opinion and in the opinion A

lot of my peers in the industry three

traffic classes is simply not enough

you're not going to have the flexibility

or the resolution to deal with the

unique traffic on your network if you're

restricting yourself to only three

classes so it can work but I'm going to

recommend that you use more here is what

I would consider the most lightweight or

the simplest class model you should

consider in a modern Network voice media

would be a class for handling real-time

audio flows

perhaps real-time video flows as well if

you have video on your network voice

signaling obviously the the signaling

the call setup and tear down those

things are important almost as important

as the media itself for voice we have

our mission critical data our

transactional data our best effort data

and our scavenger data so this is where

you and the business may have some

disagreements they may say this

um

Mainframe application is absolutely the

most important thing in the business and

you know there may be this thing like

um oh I don't know a routing protocol

you know you're going to have to

consider in your environment what takes

priority and where but we're giving you

classes and giving you flexibility in

the grouping of these things to let you

do that so Mission critical

transactional best effort and scavenger

seem to work pretty well

again and we'll show you a much detail

or much more detailed or a lot bigger

model in future slides

step 3 develop qls policies for traffic

classes here is an example of the qos

policy you know or potentially one Qs

policy anyway that we might apply to

those classes I Define so voice media

we're going to Mark the packets with a

dscp value of EF or expedited forwarding

and we're going to talk about the SCP in

the next video so don't you know

don't try to commit this out of memory

and I don't know what dscpef and the CS3

and all this is don't worry about that

we'll cover that so voice media we're

going to mark it with a dscp value of

expedited forwarding and we're going to

use low latency queuing to always

prioritize The Voice media why

is it the most important business

application maybe not is it the business

application that'll you know fall to

bits first and be noticed first and

generate the most number of phone calls

yeah probably so it's got its own

traffic class

voice signaling we're going to Mark

packets with the dscp value of CS3 and

this actually changed we used to Mark

the stuff as AF 31 if memory serves but

that has changed and what Cisco's doing

over the years so we're going to current

current recommendations Market as a CS3

Mission critical data we're gonna you

know Define some kind of minimum

bandwidth guarantee in this example

we're showing one megabit you know maybe

it's 10 Meg maybe it's 5 Meg you know

use a number appropriate to you we're

going to Mark the packets with the dscp

value of AF 31 for assured forwarding

and we're going to use class based

weighted fair queuing I love class based

weighted fair queuing you're going to

see me use this a lot for transactional

data af21 you know it's more important

than best effort it's less important

than Mission critical best effort data

you know we still may be guaranteeing

um some bandwidth but we might be

creating some caps we might say best

effort data gets a cap of 500 KB we're

gonna mark it with a dscp value of

default and use class-based weighted

Fair queuing

scavenge your data you know maybe this

is the stuff I don't really want on my

network

um you know maybe it's that

um e donkey peer-to-peer traffic maybe

it's

um I don't know it could be anything you

know whatever's not important to you or

or kind of the opposite of not important

but I don't want it on my network at all

maybe you put that into the scavenger

class and you use this to police it down

you know you mark the packets with the

dscp value of cs1 and we're going to use

red I'm sorry W red is how I should

pronounce that which is weighted random

early detection to I you know to tag

these packets is drop me first you know

in the event that we're having

congestion and need to start dropping

packets we're going to use this as a

flag to drop me first and more about you

know all this low latency queuing and

dscp and class based weighted for

queuing and red and all that more uh

coming up in future slides

additional considerations for voice

you need to understand the

characteristic of voice traffic it's

demand for bandwidth is what we would

call smooth it's not bursting in nature

packets are small the sizes are

predictable you're not going to have

these big giant you know flows of voice

all of a sudden and then drop off it's

going to be fairly smooth and

predictable voice cannot cannot cannot

tolerate delay

um it cannot tolerate excessive Jitter

so you've really got to make sure voices

moving and moving efficiently around

your network

there are standards out there and I

can't recite it by name which you know

which standard this is defines this but

no greater than 100

150 it's hard to say millisecond one-way

delay in the audio path

you start getting greater than 150

milliseconds and the user is going to

start perceiving

delay with voice and this isn't a you

know a call manager a Unity connection

kind of number you know those

applications actually have much smaller

numbers for delay but this is a

am I is my brain and my ears and you

know my mouth and everything gonna deal

with voice

um you know kind of number

you start trying to get half second you

know each way delay and it's more like

two-way radio or it's talked over you

know Roger sound like I'm talking on the

CB radio or something but uh you know

you got to keep the delay small or

you're gonna start getting that

um you know asynchronous type of

behavior and experience and that's not

what you want so keep your delays low

and we want to have less than one

percent packet loss with voice if I'm

talking and you start missing every

other syllable or every third word or

even every 10th word you're gonna lose

your mind because you're not gonna know

what I'm saying we're not going to be

able to have an effective conversation

and you're gonna freak out so less than

one percent packet loss for voice

when we deal with video consider that

it's bursty the video is sending

differential information and I'm not

going to get into all of the iframes and

P frames and B frames and all of that

but you know if you're interested in how

video transports you know certainly look

that up but video sends data sends

information as it changes it sends the

difference of information so even though

you think video oh that's that's got to

be a lot heavier than audio well yeah it

uses more bandwidth but it's it's

different than voice

um video also cannot tolerate delay

so that's something you need to consider

when designing for video support on your

Enterprise Network and finally data

TCP

which is what you know by far the vast

majority of Enterprise data is I'm not

saying that all data is TCP that

wouldn't be fair but we got a lot of TCP

out there these days with HTTP and FTP

and you know the list goes on and on

TCP can deal with

um

loss of data you know we've got

handshaking that goes on and we've got

um you know methods to re-transmit lost

data when we've detected that what we've

sent hasn't reached the destination

because the destination didn't

acknowledge it and you know blah blah

blah all this stuff's there

we're using it it doesn't help us in a

voice and video world but data can deal

so let's leverage and lean on the fact

that data can deal when creating qos

policies in our infrastructure I'll be

the last person to tell you to never

drop a packet I'm going to say if it

makes sense to drop a packet and it

protects the traffic that needs

protected the most and doesn't create

perceivable user impact then drop that

stupid packet and let it through the

next time around so these are things to

think about so we've covered the qos

policy approach we've talked about

strategies we've talked a little bit

about qos policy classes and how you

know how to treat traffic based on class

we're going to get a lot deeper into

that this was just scratching the

surface but you know that should give

you a good start so in the next video

we're going to talk about some of these

technologies that I've called out my

name what is gift serving dscp we'll

talk about differentiated Services

versus integrated services and explain a

little bit about RSVP and talk a little

bit more about you know how these actual

mechanisms work so with that I'm going

to say thanks for watching good luck

with your studying and I'll see you in

the next video

[Music]

thank you

[Music]

in this module we're going to talk about

two of the

qos methodologies

that Cisco is going to want you to

understand as part of The ccnp Voice

C voice exam curriculum

and as I talk about the differences

between integrated services and

differentiated Services I want to kind

of level set your expectation of what

you're going to see in the field

never

in the last decade that I've been doing

unified Communications have I personally

worked on a network that was leveraging

integrated services

now I do know that they exist

I do know of some clients and uh

people that I've worked with in the past

having these environments

um they're typically government in

nature

and are things that most of you will

never touch because by and large diff

serve is you know kind of rule in the

world so I don't want to say that

integrated services

doesn't exist or that nobody's using it

but its use cases are very specific and

you know it is by far in the minority so

let's get right into it and talk about

integrated and differentiated services

when we talk about the qos models there

are really three different approaches

you can pick When developing an

Enterprise qos strategy there's the best

effort delivery model this is what the

internet uses or you know what we need

to presume the internet is using anyway

and that that is that no qos is applied

to network transport we're not treating

any traffic in a specific way we're not

queuing or prioritizing all traffic has

best effort treatment and we do with it

what we can

so that's one qos approach it's kind of

the lack of quality of service treatment

if you want to consider it that

we have the integrated Services model or

int to serve

which is designed to provide absolute

guarantees of bandwidth

and that's what I want to zoom in on

absolute guarantees

and it does that by using a process of

resource reservations

and call admission control so RSVP is a

protocol

and call admission control mechanisms

now

integrated Services is not very scalable

it's highly complex and there's a lot of

configuration involved

and like I said before it is not the

predominant method in use in most of the

networks I've ever touched

differentiated services or diff serve

is kind of what rules the world at least

you know in my neck of the woods and it

provides what I'll call almost

guaranteed bandwidth but it's much more

scalable and flexible than integrated

services and qos treatment within a diff

serve environment is handled on a per

hop basis in fact we call these things

per hop behaviors but we'll get more

into that as we get through the slide

deck so diff serve is what Cisco wants

you to understand in depth because it's

what's popular but you may have some

questions on the exam relating to the

integrated Services model because it's

fair game within the curriculum so let's

go ahead and start there and let's talk

about integrated services

so integrated Services has the benefit

of offering

guaranteed bandwidth

for a flow

if the bandwidth is available because it

we reserve it before we make a call so

it gives us explicit end-to-end

controller resources

now the disadvantage to that is that

this flow-based approach doesn't scale

well

and it's very labor intensive and it

doesn't support large networks well

integrated Services uses the RSVP

protocol along with called emission

control mechanisms to provision or to

reserve bandwidth from end to end across

a network before flow is ever created

so you're getting a true guarantee and

I'm saying here that integrated Services

is not often used in modern environments

diff serve as much more widely adopted

as a technique so I'm not bashing RSVP

I'm not bashing integrated Services I'm

simply saying that it is a niche

solution deployed you know within silos

of its own so you may run into it it

works it works pretty well but it's it's

not the popular method in the industry

and that's most of what I'm going to

tell you about integrated Services now

when we get into the diff sir which is

kind of the world we're living in here

diff server differentiated Services is

highly scalable and it supports many

different classes of traffic and we're

going to get into this whole classes of

traffic thing in great detail now some

of the disadvantages to diffserve if you

want to call on that

you know if we're going to kind of get

down to the letter of the law here there

is no quote unquote absolute quality

guarantee because we're not provisioning

bandwidth across the network we don't

know what's available

uh when we're sending a call up we're

simply setting a call up hoping the

Network's healthy and allowing per hop

behaviors

to treat our traffic appropriately from

end to end multiple mechanisms are going

to come into play in order to seamlessly

deploy diff surf throughout an

Enterprise environment

I think that kind of speaks for itself

we've got a lot of places to configure

things

I've mentioned it before I'll mention it

again it's the most widely deployed qos

mechanism in Enterprise networks today

and it is handled or treatment of

traffic is handled hop by hop and we

call these per hop behaviors but more

about that in slides to come here

so how diffserv Works let's get some

terms out of the way they're things that

I'm going to reference throughout this

module diff serve treats traffic with a

common qos tag called a ba a behavior

aggregate similar traffic has a similar

Behavior aggregate

dscp or diffic code points are values

contained in the IP packet header that

is used to assign a treatment to a

packet so we're going to group things

together and we're going to detect them

we're going to classify the traffic

based on its differed diff serve code

point value more on that as we go and

PHP the per hop behavior is the behavior

that takes place on a network hop by hop

so queuing decisions packet scheduling

policing shaping Etc are all happening

these decisions are happening as the

flow is created

um you know hop by hop or one hop at a

time

so disserve Behavior if we take a look

at the example Network we've got an IP

phone this could just as easily be a PC

a couple of routers and an IP phone so

we've got multiple hops along the way

the flow of data is going to be end to

end obviously and what's going to happen

is as the packet travels from the

originating device to the first router

the first layer 3 device

um obviously when this hops on the

network we're going to classify traffic

we're going to identify it and Mark it

appropriately at the network edge with a

dscp value

as we egress the first router to the

second we're going to apply a per hop

Behavior so we're going to decide on our

queuing mechanisms our packet scheduling

our you know Etc as we egress

the next router is going to do the same

and the next router is going to do the

same so it's very unique it's very

independent of the end-to-end

infrastructure now you still need an

Enterprise qls strategy to deal with

this but it is a hop by hop decision

that's taking place

understanding dscp when we talk about

diffsert code points basically these are

numeric values and we use them to

identify a class of traffic now I'm

going to get into a really nice chart

and show you the diffserv code points

but I want you to understand what's

actually happening here we're going to

get into understanding the IP packet

header

any type of service field which is an

8-bit field and we're going to talk

about what the individual bits in the

field are used for and how this all

really works so here is an ipv4 packet

header and I've simplified it to just

focus on the type of service byte today

we're calling it the DS field

and I'll explain what I mean by today

we're calling it as we get a little

further down here as with all standards

they evolve over time so early on we

used

something called IP precedence actually

you know early in the beginning we used

something called type of service in an

ipv4 packet header later we moved into

IP precedence and finally today we're

using diffsert Code points

so the DS field and the IP header right

here

is used by diffserv formerly we called

it the toss byte but today it's called

the DS field

as things evolved or I should say in the

beginning RFC 791 introduced the

specification for the type of service

field the first three bits

were used for IP precedence so that's

these bits right here

the next three for delay

throughput and reliability so that was

IP precedence

Along Came RFC 1812

and it modified the definition of the

toss field

by removing the meaning of the last five

bits or marking them zero

um

and this is what we call you know RFC

1812 IP precedence so an evolution was

happening so now you know these bits

don't mean anything anymore

and they're all zero so what's that

leave us it leaves us these three bits

and those are going to create what we

call IP precedence values

as Evolution continued RFC 2474 replaced

the toffs field with the DS field and we

stopped looking at IP precedence and we

started looking at dscp

NECN now we'll get more into each of the

bits in this field in another slide but

I want to show you you know just at a

high level the ipv4 packet header and

the toss or the DS field

so dscp values here's what we're really

going to break it down for you this

chart is a lot of fun

um I've written different variants of

this chart over the years I'll never

memorize it although I wish I would but

I really want to show you exactly what's

happening so when we were using

ipprecedence

I told you in fact I'll Circle it here

we were using the first three bits bits

0 1 and 2.

so we had

you know values of 0 through I have to

do my math there

how high can that go

see one two three four what is that like

seven I think we can go up to seven

my binary counting is a little rusty

um

we've got three bits to play with

an IP precedence value of one

actually let's start with zero

is going to have a value of zero zero

zero in those three Fields so we're

doing a normal binary Counting

in IP precedence value of one

is going to be 0 0 1.

so we've put a A1 or a flag if you will

in the uh the first bit column

in IP precedence value of 2

was 0 1 0.

nip precedence value of 3 was 0 1 1.

an IP precedence value of four was one

zero zero

an IP precedence value of five was one

zero one a value of six was one one zero

and a value of seven was one one so you

can see how we're using just these three

bits is all we're using

so that is our IP precedence value

so that covers you know the first half

of the chart let me go ahead and uh

clear the ink here

and we'll keep going as time went on and

as we began to adopt dscp

we created dscp classes

and there's the default dscp per hop

Behavior a default class

which has a dscp binary value of zero

zero zero zero zero zero

and I want you to notice that the first

three bits

of the dscp binary value

equal

the IP precedence

binary value

because they're the same bits it's these

bits

right here these first three bits

so that's

right there

the second three bits or the

fourth fifth and sixth bit

or if you're counting from zero three

four and five here

that's the next three so together

those make the dscp value

and what's Happening Here is we've made

dscp

backwards compatible to some extent with

IP precedence

so that devices using the network that

honored the previous means of Ip

precedence could still somewhat deal

with modern qos mechanisms so we've

talked about the default class

now we're going to talk about expedited

forwarding EF down here at the bottom

whoops EF

EF is used for typically priority queued

traffic this is going to be our voice

traffic

you don't use a lot of EF traffic on a

network in fact most of the time the

vendors will recommend or sometimes

mandate that you use less than 30

percent of Link bandwidth

for EF based traffic

EF is going to be a dscp value of 46 in

decimal or a one zero one one one zero

which is going to equal an IP precedence

value of 5. so that's EF

assured forwarding has multiple classes

in fact we have four

um classes of assured forwarding

and in fact let me uh let me fix

something in my notes here because four

classes with four drop priorities is

supposed to be talking about the shirt

forwarding this is

something separate so a shirt forwarding

we have four classes with drop

priorities and the way that this works

is we've got

AF 11 12 13.

we've got AF 21 22 23.

and we've got AF 31 32 33 and 41 42 43.

so here's a class

here's a class here's a class and here's

a class

so in the AF 1 class

we have a low Drop priority so af11 this

last one is the drop priority so it is a

low Drop priority the higher the number

the higher the drop priority

af1 drop priority two

af1 drop priority three

and so on and so forth

um

the

AF 1 class I just want to back up here

contains a low Drop priority

which is here

a medium drop priority and a high drop

priority and the same thing recycles

here we've got low medium high low

medium high low medium high whoops H for

drop priority so that's our AF or our

assured forwarding classes

and assured forwarding classes are

what we do for most you know most data

and what's good about assured forwarding

is that it's going to allow us to

allocate or guarantee a certain amount

of bandwidth to the class however if

extra bandwidth is available will allow

it to access it

we'll go ahead and we'll clear the

screen again here

and let's talk about class selector we

have cs1

CS2 CS3 CS4 5 6 and 7.

the Cs classes or class selector

is giving you

bit for bit backwards compality

backwards compatibility I should say for

dscp with IP precedence

so you're getting

a way to map

to a precedence of one you're getting

away to map to a precedent so two you're

going to wait a map to a precedence of

three

so slightly different and you know

you'll find that um

voice RTP on a Cisco phone

is going to be EF so we'll say voice

RTP

and the

um

the signaling for the call is going to

be CS3 now back in the earlier days

Cisco used AF 31 so we're just going to

do this we'll save voice

signal Lane but today you know it's it's

up here at CS3

um

the biggest thing I want you to

understand when looking at this drawing

or looking at this chart

is that

there is a method to the madness we're

not doing

um anything all that crazy here we're

simply looking at the bit positions

these three for IP precedence and then

we're going to assume these are all zero

these

six for dscp

and you're good to go so

Network endpoints routers Etc that were

compatible with IP precedence to some

extent can still function and deal with

packets that are tagged using dscp so

what I'm showing you here is just a

chart don't freak it out you know about

it don't try to memorize it you know

it's an important absolutely

what I want you to understand what's

happening more than I want you to

remember these exact numbers know that

there are four assured forwarding

classes know that you've got a low

medium and high drop priority know that

there are seven

um CS classes know that there's EF if

you've got that

and as long as you understand that the

first

three columns yeah there's no way I'm

gonna keep this line straight but the

first three columns in the dscp binary

value are going to equal VIP precedence

binary value then you should be good to

go so we'll move on to the next line I

know that was a lot to cover

and let's talk about the diffserve qos

mechanisms we have got classification

and marking congestion management

congestion avoidance policing and

shaping and Link efficiency mechanisms

so all these different mechanisms

available to us within the diff service

qos model

classification and marking we're going

to classify traffic at the input or the

Ingress interface on our device

we can look at its dscp value its IP

precedence value and Source or

destination address and we can do this

classification either via the mqc the

modular qos CLI which is what you're

going to see me do in the next uh you

know the next section we'll actually do

some Qs configuration

it can use n bar or policy-based routing

and classify based on an ACL

congestion management and I'm not

talking about Sudafed there congestion

management gives us queuing algorithms

and this lets us deal with congestion as

it's happening so fifo first in first

out priority queuing custom queuing

class-based weighted fair queuing and

low latency queuing are all available to

us I want you to know that low latency

queuing is the preferred method by Cisco

and is a hybrid It's a combination of

priority queuing and class-based

weighted fair queuing so basically we're

going to strict priority queue voice RTP

and then we're going to wait at Fair

queue everything else based on class so

it's very flexible and it works really

really well

congestion avoidance

we've got congestion avoidance

techniques so we can Implement on egress

or output interfaces

things like weighted random early

detection

wret is Cisco proprietary

and it's useful

for dealing with TCP based flows and

what wred is going to do

is it's going to drop lower priority

packets instead of dropping higher

priority packets so it's got some some

evaluation and decision making going on

here

that can help us deal with

um you know there were a stop from

happening I should say because this is

all about congestion avoidance uh link

congestion

policing is good for controlling burst

and it helps to make sure that the types

of applications that need bandwidth can

get their bandwidth policing is going to

drop or Mark packets once predefined

limits have been reached and you're

going to see policing used you know in

service provider environments a lot I've

done it in the Enterprise but I don't do

it often in the Enterprise so the vast

majority of what's happening in the

Enterprise is classification marking and

then you know that's pretty much it I'm

not doing a whole lot of policing

I'm also not doing a lot of shaping but

you know from time to time I do so

shaping is used on egress interfaces and

it's allowing us to structure

what and how gets placed on an egress

link it's going to keep low speed

lengths or help keep low speed links

from being saturated by you know flood

of traffic we've got available to us

both generic traffic shaping and frame

relay traffic shaping so you'll want to

know those for the exam

compression again something I'm not

doing a lot in modern Enterprise

networks at least not here in the United

States because our links are relatively

fast

two types of compression are available

you've got header compression and

payload compression header compression

or compressed RTP is going to reduce the

size of the RTP UDP and IP headers

however it's going to come at a cost of

CPU overhead and payload compression is

going to give us layer 2 frame

compression using either the stacker or

predictor methods again these are things

that you want to use on low speed

lengths only because they do make your

router work quite a bit harder and the

number that I like to throw out there

and I don't remember where I read it was

don't even consider using this stuff

unless you're on links slower than 768k

So you you're half a T1

lfi link fragmentation and interleaving

is also a technology that's available to

you and you can use it on slow hand

links to fragment larger data packets

and interleave smaller voice packets

within where you know the areas there's

larger date of tackets would have been

it would have been sitting so

this is a lot I know we've been kind of

drinking by the fire hose in this module

but this should give you the

fundamentals of integrated services and

more importantly differentiated services

and when we get into the next set of

videos or the next uh the next section

we were actually doing qos configuration

you'll understand what I'm referring to

when I talk about classifying traffic

and marking traffic and and queuing

traffic and you know weighted Fair

queuing versus llq versus priority

queuing

um you'll probably want to go back and

watch this video a second time like I

said drinking from the fire hose you

know that chart alone on dscp values and

how they map to

um you know some of the other uh you

know type of service byte fields and uh

you know IP precedence Fields you know

can be a bit daunting especially if

you're trying to memorize it so spend

the time get a good understanding of

what's Happening you know don't sweat

the details right now you know those are

things you can go back to prior to your

exam and you know try the cram in your

brain

but right now I want you to understand

it I want you to get a good feel for

what's happening behind the scenes so

that's it for this video I know it's

been a little bit of a long one I really

try not to make them this long but you

know I feel that you know for dscp and

integrated services or diff service and

integrated Services it's best to just

kind of get it out there on the table

and talk about it all at once

um rewind this one do it two or three

more times it should hurt a lot less the

second and third time through and I'll

see you in the next video where we're

going to be talking about a benchmark or

a baseline qos model that is being

pushed these days and I'm not sure

exactly where it originated

um I've seen like I mentioned before

I've seen temps to Getty use it in his

Enterprise qos book I know Cisco press

is using it in some other books I've

seen other engineers in the field using

it so we're not going to spend tons and

tons of time on it it's only a couple of

slides but I do want to show you that

Baseline that I like to use in my own

environments and it'll give you

something to build on so with that

thanks for watching good studying and

I'll see you in the next video

[Music]

foreign

I know after that last video your head's

probably spinning just a little bit

hopefully you've taken my

recommendations and rewound and watched

it a couple more times if you haven't

stopped now go back and watch that again

it really is not

terribly complex but it's it's so much

information to take in at once that you

just kind of need to spread it out a

little bit this one this video here is

wrapping up the fundamental knowledge

you'll need of qos for the C voice exam

from a theory perspective

and it's only one slide and it's not

going to take a lot of time so this is a

good wind down exercise or wind down

exercise to kind of decompress a little

bit and think about qos so let's get

right into it and this is the 11 class

Cisco Baseline qos model I've seen

variations on this theme for years and I

really like it I worked at a large uh

you know multi-billion dollar a year

Healthcare organization Healthcare

Organization for a number of years I was

a network architect and we implemented a

very close approximation to this 11

class Baseline qos model and for like

eight years used it and it just it grew

and changed and adapted with us very

very well

and the beauty of an 11 class modeler of

this 11 class modeler really the class

models in general is you can add to them

and subtract from them at any time you

know if you design it flexible enough in

the beginning

that um you leave yourself some Elbow

Room you can modify and tweak things as

necessary

what I want to show you here is

one concept this is a very detailed

methodology that allows us to in my

opinion

clearly other people's as well maximize

our use of different dscp values

to kind of get the most

capability and flexibility out of how we

tag our traffic so let's just kind of

start at the top which is going to be in

fact the order you see this in is going

to be kind of the highest priority to

the lowest priority

and you're going to see that you know

things like Mission critical data

um you know are kind of a little lower

than you might think that they need to

be and you know web traffic General data

you know normal stuff's almost clear at

the bottom and your business you know

your management if they were to look at

this chart

would not understand why they would not

get

that

you know sometimes technology needs for

things to operate properly

come before the warm and fuzzy but I'll

let you manage that on your own that's

another challenge but anyway starting at

the top the most important traffic on

your network

and what needs to be prioritized is the

CS6 is routing protocols and network

controls so the stuff that keeps your

network up

is more important

than the applications that the network

is servicing obviously

the network you know the apps can't

override these things that keep the

network running so it really needs to be

number one above any and all

applications from a qos perspective

Cisco recommends a CS3 layer 3 per hop

Behavior with rate based queuing and

random early detection

voice RTP

a close second that's our Bearer traffic

or the audio content

we want to do that as an expedited

forwarding which is going to be a call

admission Control Plus strict priority

queuing so we're going to guarantee

bandwidth for voice

dial tone just has to work so that's

that's why it fits in the way that it

does video RTP the bear traffic for

video is going to be next with an AF 41

which is CAC plus rate based queuing

plus weighted random early detection

streaming video very similar that's our

streaming media content but not

um you know not video conferencing would

you know I'm when I see video I'm

thinking video conferencing two-way you

know streaming video would be a one-way

thing

and that's going to be a CS4

again CAC plus rate based queuing plus

weighted random early detection

Mission critical data comes next that's

the critical Enterprise data apps

so your business is number one most

important we've got to have this to do

the business app I don't care what they

say fits here you know at number five AF

31 rate based queuing plus weighted

random early detection

call signaling that's your call set up

tear down and any other kind of you know

necessary Communications to keep calls

running for voice and video is going to

be a CS3 and rate based queuing plus

random early detection the reason we're

using red instead of w red here and in

fact you'll see the same thing happen

and now with network management is that

weighted random early detection is for

TCP

and uh you know these are typically UDP

signaling flows so that's why you'll see

red down here it's not an absolute

statement but that's kind of a

generalization

uh transactional data things like

database apps Mainframe apps you know

your telnet sessions I want to put them

in here af21 rate based QA plus weighted

random early detection

Network management things like SNMP

we're going to give a CS2 rate base

queuing plus red

most of the rest of the data on our

Network that we're not going to classify

and really you know like 70 of the stuff

really ought to fit here guys so try not

to go crazy on what you classify up

above

um

bulk data you know that's web traffic

and generally everything else

af11 rate based queuing plus weighted

random early detection

scavenger traffic class this is our

nuisance traffic this is our

um oh well what I want to say you know e

donkey peer-to-peer traffic and things

along that along those lines you know

stuff that we really don't want on our

Network and we definitely don't want to

give special treatment that is scavenger

class traffic and then finally best

effort that's the default class that's

everything else

um you know we're going to give a

bandwidth guarantee

um with rate base queuing plus red so

the 11 class model is going to give you

a lot of capabilities and a lot of

flexibility now keep in mind you don't

absolutely have to do this

If this just makes your head spin and

you say dang it Josh I don't want a

baseline model I don't want 11 classes

you're on drugs then I'll say how about

an eight class model

and you know if you agree to that then

we're just going to combine the voice

and video RTP whoops I'm clicking ahead

of myself the voice and video RTP into

one

we will combine

the

um

routing and network management into one

and we'll combine the mission critical

and transactional data into one and

we'll reduce the number of classes you

have to deal with

I gotta tell you there's not really that

much complexity that you're that you're

losing you know this stuff isn't that

hard to set up but you know if you can't

handle 11 classes take it down to eight

if you can't handle eight take it down

to five and really this is this is the

smallest that I would ever recommend you

do you know there are ways to take it

down to a three crap three class model

but I don't really subscribe to that so

if you wanted a five class model you

could have a real-time class that's your

voice and video

you'd have call signaling in a separate

class you'd have critical data which

would be your network control your

critical data your bulk data and then

you'd have a best effort and then you'd

have a scavenger so that would get you

down to five classes so

ways to reduce it yes

um but let me twist your arm a little

bit go for the 11 class model there's no

real reason not to it's a couple more

lines of code it's not like it's gonna

you know confuse your you know to the

point of not being able to deal with it

give yourself the flexibility design it

in now so that it scales and works well

for you for years to come so

that's probably more time than I really

should have even spent on the Baseline

qos model but I find that it's helpful

to give you a reference and you know

just like every book I've ever seen on

qos tends to give you a baseline

reference model so I figure we should

probably follow suit and do it here as

well and really give you an idea how

these things stack up so with that I

want to say thanks for watching this

wraps up our video section on the theory

behind qos in the next section we're

going to be covering configuration of

qos on Cisco routers and setting up our

Network and really a little bit of

switching for that matter

and setting up our Enterprise to handle

all these qos methods that we've talked

about and and actually doing the

implementation and that is going to be

the end of the C voice course so you're

really close we're almost there thanks

for hanging in there with me I know it's

been a lot to uh to deal with but I'm

hoping you're having as much fun as I'm

having and with that I'm going to say

thanks for watching good studying and

I'll see you in the next video

[Music]

foreign

welcome to module 34. we're going to go

through some demonstrations of creating

qos configuration on a Cisco router and

specifically we're going to focus on

classification and marking of traffic

now there's a lot of different

um

different things that play into

Enterprise qos there's land-based qos

and considerations for the switched

environment and we're not going to dig

crazy deep into that

there's Wan based cubomass where we're

trying to prioritize traffic

going from a high speed Network to a low

speed Network we're going to focus on

that in great detail

and kind of everything in between

literally so let's kind of start at the

beginning I'm going to give you a quick

little sketch of what's going on in the

network and then we're going to jump

into our 2811 router here and actually

do some qos configuration

now I'm going to sketch a little Network

for you I'm going to show you Josh's IP

phone

it's connected

to a switch a layer two

switch

and Josh's layer 2 switch

is connected to a

Cisco router

and that router

has a T1 line

that's part of an ipwan

that T1 line happens to go

to another router

on that router

is another layer two switch

with another

Cisco IP phone

when the phone creates an RTP packet

that RTP packet

that is sent to the switch

is going to have a layer 2 and a layer 3

qos marking

the layer 2 marking

is what we call a class of service

marking

and we're not going to get heavy into

layer 2 qos within the C voice exam it's

something you'll do now and again

I would recommend you deploy it but by

and large most of your qos exercises are

going to be from a routing perspective

so

layer 3 markings are going to be intact

on this packet as well and that is going

to be a DS CP value

within the toss

byte of the IP packet header

so Layer Two markings layer three

markings and keep in mind

back from your CCNA days that a layer 2

frame

contains

a layer 3 packet

so this switch

is not gonna

monkey with this dscp marking at all

it's a layer two switch

it doesn't look at it doesn't see it

doesn't care that layer 3 marking

the

switch is going to pass the packet to

the router so the router is going to see

an incoming packet

that it doesn't know what this thing is

this is Noah's voice it just knows that

it has a layer 3 dscp marking of e f

expedited forwarding

so we're going to say mark

ing of EF so that packet comes in with a

marking of EF

we are going to create

classes

inside of class Maps on the router

that identify the traffic and allow us

to group things together so we're going

to have like VoIP

RTP

VoIP control

we'll have a class called class default

you know for everything else and you'll

continue to build on this

as you deploy your qos model hopefully

that 11 class Baseline model that we

showed you

as traffic hits this egress interface

and the same would be true for packets

coming in the other direction

as they hit that interface

this is where we need qos to do its

magic that's our congestion point you

know I've got this you know 10 100

1000 base T Network out here running it

potentially gigabit or faster speeds and

this little

1.544 megabit T1 interface

so where do you think dropping packets

is going to happen you got that right

it's at the low speed egress interface

so we're going to talk about how to

configure the class Maps

how to configure the policy maps on the

router and how to deal with the

classification and marking of traffic so

let me go ahead and clear this screen

here

and we will jump over to our handy dandy

2811 router

and this router has absolutely no qos

configuration on it whatsoever

so let's walk through some basic class

map configuration

I'm going to say class map

and then I've got two options here I can

either go match all or match any it's as

obvious as it sounds if I match any if

any parameter in here matches my packet

then I am in this class

if I use match all then I'm going to

have to to match you know every

condition within this so I'm going to

almost all the time I'm going to say

match any

and we're going to create a name for the

class and like I said before VoIP RTP

now that I've created a class map

I'm going to configure

the matching rules so we're going to say

match

um let's say dscp

and then here's our options I want to

show you what's available to you EF

I could also use the decimal value if I

wanted to

I could also say match IP precedence

five that would be the same as the

dscpef but perhaps my legacy application

didn't tag dscp it tagged IP precedence

um

perhaps I've got traffic coming from a

VLAN and I've actually got an access

list I've created we'll say match access

group 101 that that's a simple IP access

list with a source network of 10 I think

ten one zero zero slash 16 going

anywhere so any traffic from that subnet

it's going to be VoIP RTP you know it

just is because we said so

so we've configured our match rules

and let's go ahead and create another

class exit we'll say class map VoIP

control and actually

match any VoIP control we'll say match

um I don't remember off the top of my

head with the IP presence value so we'll

just go dscp CS3 and AF 31 we'll assume

for a moment that those are how my

packets are being tagged

so I've created two class maps and in

fact if I do a show class map you're

gonna see

the class default which is there by

default which says match any

you'll see the vleip RTP that I created

with a match dscp EF and you'll see that

it for convenience shows you the decimal

value

it has a match IP precedence 5 and then

a match for that ACL

and then we've got our VoIP control that

we've configured so we've got some class

Maps going on now that I've configured

and classified my traffic

I need to decide

what to do

on how to treat it on my egress

interfaces now there's lots of things I

can do

with

um

you know with qos on these egress

interfaces I'm going to start off simple

just by showing you how to remark the

traffic let's say that I matched this

ACL and it didn't have a dscp value but

every other router Downstream for me is

going to look at that dscp field so I

really need to make sure something's in

there so I'm going to put something in

there

so we're going to create a policy map

it's going to be policy map and I'm just

going to call it qos

and then we're going to say class VoIP

RTP

and we're going to say set

and you can see all the things I can set

dscp

EF so now

everything that matched the class

of VoIP RTP

it's going to be re-tagged with ef and

I'm going to use EF through the rest of

my network so I've I've put some other

detection mechanisms in there in fact

I'm even going to go back and do one

more just because I want to show it to

you because it's cool

Chevron pipe section

class map

and the one I'm going to do is I'm going

to say class map match any VoIP RTP I'm

going to say match

protocol

RTP that's what we call nbar network

based application recognition it is a

little more CPU intensive but it's a

really cool way to to deal with qos

and how to detect packets so we've got

that so let me just sum it up again here

show around pipe section class map so

we've got our class maps for Vape

control and vape RTP and a bunch of

different match rules Show run pipe

section policy

map

and you'll see we've got a policy map

qos

a class VoIP RTP and I'm doing a set

dscp EF

um

that's classification and marking right

there

now to attach this policy to an

interface

I don't think I actually have a T1 in

here that's configured for data let's

see what I've got on here show IP and

brief

it doesn't really matter I don't have

one for data

but fast internet zero one is not in use

so we'll use that let's pretend that

that's my egress interface I'm going to

say interface this would be exactly the

same on a D1 and fa01

and we will say

policy

I'm sorry

service policy output qos

so I've applied the policy map using

service policy

and if you think about it this way

classes

get configured in a class map

class Maps get configured in a policy

map

policy Maps get assigned to egress

interfaces

as a service policy

so right there we've actually configured

basic qos

and

that's going to handle

the basic layer three

functionality for classification and

marking now we haven't actually

configured

any

queuing behaviors

or any kind of policing and shaping or

anything like that

we will be doing that I'm going to go

back to these service policies and we'll

perform that configuration in a later

slide so for now let's keep it simple

classification of traffic and marking of

traffic

now there's one more thing I want to

talk about

and I I hinted to it lightly

within the

um you know the PowerPoint as we were

getting started and that was about Layer

Two qos

Layer Two

qos is a little different than layer

three so first off IP headers are

preserved end to end across the network

because they're inside layer two frames

however

layer two frames are not preserved

they're recreated from switch to switch

to switch as packets work their way

through your network

to provide

for

true end-to-end

continuity of your qos tags

at some point you really need to map

layer 2 tags

to layer three mappings

and

like I said you know the biggest

point in your network where you need to

deal with

um qos is definitely the way in the

congestion points

but you will get into some basic

layer 2 qos

I want you to understand in fact we're

going to jump back to PowerPoint for a

second I want you to understand the

concept of the trust boundary so in a

network we'll go ahead and show you it's

real similar to what I drew before phone

switch

router

you know T1 line

to a Wan

T1 line to a router

and let's make this network a little

more complicated

to a switch

to a switch

to a switch

to a phone

now

when the phone

tags

the

traffic

I said it's going to do it at layer 2

and layer three

so you've got class of service and dscp

if the switch

has qos enabled

we have to understand

Its Behavior

with the layer 2 mappings

so let's talk for a second about trust

boundaries

you want to Define a trust boundary on

your network

for qos mapping

as early as you can

when a packet or frame

begins to travel across your network so

we want to Define this truss boundary

out here at the poor

where the iPhone connects so that's our

trust boundary

if

We Trust

the markings at Layer Two

then if qos is configured

we will not strip the markings

if we do not trust the port

then we will strip the markings

and Market is a preference of zero

now

it's only going to matter

as it propagates through the switched

portion of the network

as soon as it hits the router

this router is not looking at that layer

two frame it's looking at the layer 3

pack and inside of it which still has

that dscp EF in it so you know

everything across the land is going to

work but this poses a concern out here

at the other site so when this phone is

originating traffic

let's say that this is the trust

boundary here actually that's not it

this port's the trust boundary

that Qs is enabled and we are getting a

class of service mapping we want to make

sure that from here to here and from

here to here

and ultimately from here to here there's

no reason not to that that marking

remains intact particularly between the

switches because if I experience

congestion on the switched Lan I want to

be able to leverage my qos protection

now I'm not going to get into all the

layer 2 configuration right now but I

want you to understand this trust

boundary so assume as we continue

through this example

that we have defined that trust boundary

and in fact I'm going to show you how to

define that trust boundary on a switch

so stand by one second we'll bring the

3550 up

so here's our 3550 switch and I'm going

to show you a real basic how to define

that trust boundary

we're going to go oops let me click in

the right window here conficti

and globally I want to enable MLS qos

so we've we've done that we've enabled

MLS qos

I'm then going to go to an interface or

perhaps all of them into range fa01

through you know 23.

I'm going to say MLS qls

Trust

and I can tell it

what to trust

I can say

Trust

the packet class of service or the frame

class of service

which I'm going to do

and I could also say

MLS qls trust device

Cisco phone

so

if this switchboard

detects

that a Cisco phone is connected

it's automatically going to trust the

layer 2 qos tags

pretty cool huh now

when you get into layer 3 switching

you're going to be a little more

concerned about this

and

you know it's still something you have

to think about the layer two world but

you know by and large I'm doing the sun

layer 3 switches but we want to make

sure that we have

um class of service

to dscp mappings

because I want to make sure packets

leaving the switch or frames leaving the

switch

are tagged properly

now I want to clarify something I said

earlier

because

it applies conditionally and I think I

may have said it in such a way that it

sounded like it applied globally

when a phone marks layer 3

dscp values on a packet

I said that those

markings are preserved end to end

now that's true

in a network

when we don't have switches

running qos and trying to manipulate

values

if you choose not to implement lan-based

qos on your switched environment qos is

off on your switches

those layer 3 markings will be intact as

they reach your router if you do turn on

land-based qos

the switch contains a dscp to cause

mapping table

and what's going to happen is it's going

to rewrite based on the trust it's going

to rewrite the class of service value in

the dscp value

based on the math now there are default

Maps

or you can configure the maps

and

you need to understand the cause to dscp

mapping

so because a phone is mapping a dscp

value 46 the EF

and the cost value of 5.

that cost value of 5 by default is going

to get remapped to some dscp value

um

your switch may not be mapping that

dscp5

to a DS to a I'm sorry that cost five to

a dscp of 40.

if it's not you know because I'm sorry

246.

if it's not that's not ideal because

when it comes out well that's not EF

anymore that's dscp of 40.

so you can actually manipulate these

dscp to cost mapping

within the switch and let's take a look

at the switch here and zoom in a little

bit more into how this works

there's a command we can run show MLS

qls qos Maps cost dscp

and this is going to show us the default

mapping for our switch so in this

example this switch is going to take

packets with a cause of zero and map

them to a dscp of zero cause a one and

eight and let's jump straight up to five

because that's how our packet's going to

be tagged and yeah sure enough it's

trying to rewrite a cost value of five

to a dscp value of 40. well that's not

what we want we want these things to be

46. so we actually need to manipulate

the map a little bit we can do that

really easy

with a command called MLS costs

map

hang on a second

MLS qos

map

cost

dscp

whoops got to spell things right cost

gscp and it'll let us Define the values

so let's say zero

for a cost of zero

10 for a causal one

we'll say 18 for a cause of two 26 for a

cost of three

34 for a cause of four

now here's where we're customizing a

little bit 46 for a cause of 5. 48 for a

cause of 6 and 56 for a cause of seven

so now we've redone that map so check

this out

so now when I look at the map

a COS of five

is now going to map to a 46 which is the

SCP EF so you want to make sure you're

doing this cause the dscp mapping on

your switch

closest to the trust boundary so that

the packets are being remarked

appropriately

in fact it's not a bad idea to go ahead

and do this you know on all the switches

in your environment from a consistency

perspective

because after all traffic can originate

lots of places and be manipulated lots

of places so you want to make sure that

this is intact end to end so we're good

there

so that's really all that I care about

showing you on a switch so back to our

router

you know we've done the classification

we've we've shown you how to do the

marking we've shown you how to apply the

service policies and really that's about

all we need to do on this video we're

going to go through some other examples

or I'm sorry we're going to continue to

build on this example in some other

videos and we'll talk about things like

the

you know the um interleaving and

fragmentation we'll talk about the QA

and what we can do with that

um you know go through some examples on

configuring shaping and policing low

latency queuing and those kinds of

things but for now classification and

marking you've kind of got it covered so

again on the router Show run pipe

section

class map whoops back up here this is

kind of where we're starting we've got

VoIP control and vape RTP

Show run pipe section policy map we'll

show you the policy map qos where we've

gone ahead and set a dscp value for a

particular class

and in a you know another video we'll uh

go through the

queuing and other Associated

configuration so I'm babbling it's been

a long video I know we're again drinking

from the fire hose a little bit but

having a lot of fun doing it so thanks

for watching good studying rewind this

one if it doesn't make sense there's a

lot of good stuff in here and we're

we're feeding It Fast and Furious so

I'll see you in the next video and talk

to you soon

[Music]

thank you

[Music]

in this module we're going to talk about

link efficiency mechanisms and this is

going to be most applicable to those of

you in a part of the world where low

speed you know sub T1 or fractional T1

services are still in uh you know a lot

of use so you know those of you in the

US who are using you know multi-megan

gigabit interfaces and mpls with gobs of

bandwidth and Metro ethernet you're not

going to use these kind of things but

slow lengths you know we need to squeeze

as much out of them as we can so let's

talk about link fragmentation and

interleaving so what in the heck is link

fragmentation and interleaving well this

is simple picture you get a slow circuit

a 56k line and you're throwing a 1 000

byte packet on The Wire you're going to

be waiting a few milliseconds or you

know more than a few milliseconds for

that packet to be put on the wire

and what's happening with your voice

it's waiting that's a problem so what

link fragmentation in interleaving is

going to do or lfi is it's going to

allow us to chop up those big data

packets into smaller ones it's going to

allow us

to then stuff some voice in between so

if I wanted to give you a quick little

sketch example here let's go ahead and

pull the pen up here and uh I'll show

you kind of what's going on

and that'll work

um if I had you know Big Data packet

and

I had a little bitty voice packet behind

it

and I'm sending traffic

and this is my interface on the router

I gotta wait for that whole data packet

to cross the link before that itty bitty

voice pack can go and then I might have

another huge data packet after it so you

know voice is not doing necessarily the

most Optimum thing what link

fragmentation and interleaving is going

to let me do is it's going to let me

take this this data packet and chop it

up

so data voice

data voice

data voice

and send it like that so I'm able to

throw this voice packets in there

help get them out the wire faster and

the data is just being sent in multiple

chunks so that's kind of it right there

go ahead and clear the screen here

and walk through a little bit more so I

talked about it already you know we're

breaking up to Big Data packets and then

leaving Small Voice packets with them

the net result of this is going to be a

gain

of performance for you it's going to

help you reduce the packet delay and

Jitter

there are two lfi types we can use if

you're doing multi-link PPP you can use

MLP interleaving and if you're doing

frame relay you can use frf12

MLP is you know by far the most widely

used frame relay is all but gone in this

part of the world I know again still

exists other places so I'm not but uh

you know we're mostly beyond the frame

relay days here

um we're going to show you in this chart

that link speed really really really

comes into play here if you're greater

than two megabit Cisco recommends you

don't even you know think about link

fragmentation interleaving or compressed

RTP just leave it alone if you're on a

really small circuit less than 768k you

know use lfi Cisco recommends that

you're going to use compressed RTP and

you should avoid using video on these

links because you just really don't have

the bandwidth anyway and from 768k to 2

megabit you know it's kind of up to you

lfi is not necessary use it if you want

and compressed RTP also it's not

necessary use it if you want

this chart is really going to show you

you know the benefit not the benefit but

you know how bad these serialization

delays can be for a given packet size on

a given link speed a 56k line with a 64

byte packet it's going to take roughly

nine milliseconds to serialize that

packet to do everything necessary to put

that bit on The Wire

you increase that packet size to 10 24

and now I'm waiting 144 milliseconds

holy cow that's a long time

um the voice is almost not even possible

anymore on those kind of links when you

get into those metrics here we get a

stable of this 150 milliseconds

um you know a pen link speed to 512k and

then the difference is 1 to 16

milliseconds you know you get into one

Meg circuits t1's and bigger this just

the problem just disappears it's not a

problem anymore

but these really slow links

um low speed lengths definitely can see

some gains from this now I've got a

router here and I've got a really basic

frf12 configuration setup on it and I

wanted to show you exactly what it takes

to do this this isn't a live frame relay

link so I'm not going to give you you

know any real-time statistics or

anything like that but I do want to show

you the commands the first thing you're

going to do when configuring fr12

obviously you're going to need to have a

working

frame relay configuration

and what I've got on this router is I've

got a Serial interface

and we'll show it to you right here

we've got a Serial interface configured

for frame relay

and we've got a 0.7 interface with the

interface Del C 101 now I've created a

map class

called frame right here math class frame

relay frame

and I've defined my frame relay traffic

shaping parameters my commit an

information rate and my burst class and

my uh my fragment size and it's set up

for fair queuing all I really have to do

is I go to this serial interface

and I reference that frame relay map

class

from within the sub interface

and we're going to be then able to

leverage the frame relay or the frf12

link fragmentation and interleaving and

you know we're defining our fragment

size down in these map classes now you

won't be able to see any output because

I'm going to actually have this thing

running but to show frame relay fragment

is the command that you could use to get

some statistical information on the

performance of this and that's really

all I'm going to talk about with frf12

we don't need to get into full-blown

examples because it's not that big of a

topic and it's not something you're

likely going to run into I do want to

talk quite a bit about multi-linked PPP

in fact give me just a second and we're

going to back some of this frame relay

configuration out and program this for

mlpp so interface serial zero zero zero

colon zero end cap PPP

um let's see here that should have got

rid of my frame relay sub interface if

it didn't then I'll get rid of it by

hand

let's let's see if it did yeah it did

okay so zero zero zero colon zero

encapsulation PPP

I'm going to first create a multi-link

interface you know be multi-link you

know you can use this to aggregate

circuits so interface multi-link one

and then we're going to say

um

multi-link

enter

now I'm going to say PPP multi-link

interleave

and hit enter and that is enabled

interleavening of fragments

now I'm going to go

to the um actually let me let me do a

show run on that real quick

and make sure it's all good before I

continue

so there's our pan where's my multi-link

there it is pp3 multi-link group one PPP

multi-link and multi-link interleaving

okay cool I'm going to go to my serial

interface configure t

and serial zero zero zero call one zero

and whoops what did I do wrong oh

spelled it wrong

and

um have a great day here with the

keyboard perfect I'm going to say PPP

multi-link

and then group one so you know this is

basic multi-link configuration

if I want to I can specify the maximum

desired fragment delay

for the interleave multi-link connection

and I'll do that under the multi-link

interface itself so ENT multi-link one

and I'll say PPP multi-link

fragment delay and then I've got between

0 and 1 000 milliseconds you know so I

can say you know five milliseconds or 10

or 15 or whatever I want

and uh if you want to get statistical

information and some stats on how this

is working show interfaces multi-link

let's see oh I need an interface number

one it's down so you're not going to get

a whole lot but you can get information

about the performance of the multi-link

interface so really those are the two

lfi methods that Cisco's going to expect

you to use and understand for the C

voice exam

if you want to get into header

compression and do things like that it's

basically class map stuff

so not a whole lot

um you know a challenge there you know

you've done class Maps before so if I

wanted to go in here we'll do an example

here config t

class map and we'll just call it uh

v o i p

and then we'll say match protocol RTP so

we're using n bar here

and then we want to create a policy map

and we'll just call it um

you know my VoIP

policy

and then we'll go ahead and do

compression

no why is it not letting me do that

oh the class

VoIP now we'll be able to do compression

compression header IP

and then you can go RTP or TCP depending

what you're wanting to do so you know

RTP

and then you'll use this just like you

would a qos policy so if I wanted to go

like um

serial zero zero zero colon zero you

know um

service

policy output and then you know VoIP

what did I call it did I call it VoIP

let's see my VoIP policy

and yeah I didn't set up IP on the

interface but you get the point

um it's just complaining because there's

no IP address

but really that's it I mean you know lfi

is pretty straightforward

um you know header compression is pretty

straightforward again low speed links

only you're not going to be doing these

things on you know

ds3s or 10 Meg Metro ethernets or mpls

connections you know keep this you know

for those sub T1 you know less than 768k

kind of connections but I know I've kind

of babbled a little bit on this I've

honestly shown you more than you really

need to understand for the C voice exam

as far as frf 12 and the MLP goes but

hopefully this gives you a little

practical Baseline understanding of

what's going on behind the scenes and

and how lfi can work you know and when

you should use it but again rare

occurrences low bandwidth links only so

I'm starting to repeat myself I'm sure

it's getting boring

um we're really close to wrapping this

up we've got a really really interesting

section coming up next in fact this is a

lot of the meat and potatoes on qos and

this is all about queuing and traffic

conditioning so I'll see you in the next

video where the rubber is going to meet

the road on building these qos policies

and I'll talk about queuing the real

magic of qos so see you in the next

video good luck with your studying

[Music]

in this module we're going to be getting

into

configuration of queuing

and you know really this is where qos

kind of has the rubber meets the road if

you want to think of it that way

and we're going to talk Basics first

about queuing methods and how queuing

really works and then we're going to get

into the specifics of class-based

weighted fair queuing and low latency

queuing so first let's talk about

commune methods

we have strict priority queuing where we

reserve bandwidth for an application and

the app can use it nothing else can

touch it it's not the most efficient use

of bandwidth so we don't do a lot of

strict priority queuing these days we

have class-based weighted Fair queuing

where we're going to create user-defined

traffic classes and then we're going to

apply queuing policies to those

individual traffic classes basically

we're going to have a queue for each

class and different treatment rules and

then we're going to have low latency

queuing which is a way of using

class-based weighted fair queuing but

still assigning a strict priority queue

type of behavior to traffic that can

benefit from it so it's an interesting

combination of different qos

Technologies so together

leveraging class best class-based

weighted Fair queuing as part of an llq

implementation and really you can just

call it llq is you know the most popular

and widely deployed qos method used in

Cisco networks today queuing

is going to need to be performed at the

points of congestion on a network and

we'll talk about a little bit about how

that works as we go on so let's talk

about queuing Basics first off there are

two types of cues so we've got Hardware

cues what we call the txq

and Hardware cues are first and first

out period packets in a hardware queue

cannot be reordered and typically a

hardware queue is only you know a few

packets in size

and then we have software cues software

cues this is where the real magic of qos

happens

um they're used for scheduling the

packets into the hardware queue so

they're in front of the hardware queues

this is an important piece of

information that sometimes gets lost

when talking about IP quality of service

only implemented software user only

implemented when the interface is

congested and what I mean by that is if

the hardware queue doesn't have anything

in it

if a packet arrives on the router the

router is going to place the bits on the

wire and the data is going to be sent

only only only when the hardware queue

um

is full you know when the hardware queue

has data in it is software queuing going

to be used

within a software queue we can reorder

packets that's kind of the you know the

concept of what we're doing here with

the modular qos CLI and you know the

whole qos architecture that we're going

to talk about and that's going to be

manipulating the software queue

to schedule packets into the hardware

queue how we want them to be scheduled

into the hard work year let's get into

configuring basic llq and what I'm going

to do here is use the same 2011 router

we've been putting through a fair amount

of exercise here lately and we're going

to go through

um

basic llq configuration so we're going

to be doing class-based weighted Fair

command along with some priority queuing

so when you're configuring llq you know

the first thing you're going to do in

fact Let Me Show You Chevron pipe

section class map we Define class Maps

early on whoops and we defined our match

rules let me grab in the right place

here you know when we were looking for

packets based on their dscp tags Etc or

you know we were doing n-bar or you know

whatever so we've got a couple of

different

um you know a different uh you know

methods in fact let me get rid of that

that class map match all VoIP that's

something that's not related to this

discussion so let's just simplify this a

little bit no class math match all like

so we've got our

VoIP control and VOIP RTP let's focus on

that so now that they've got these

classes we are going to configure

within our policy Maps

how to treat the traffic and if I do a

show run par section policy map you'll

see the policy map that we already

created

and you know this is where we were

setting some dscp values if we needed to

do that I'm going to go ahead and go

into that policymap qos

and I want to talk about

on a per class basis

what do I want to do so let's go into

class VoIP RTP so VoIP RTP is our media

now we said before that the media the

audio

needs to be treated in a priority queue

we're going to do that using a command

called priority in fact there are two

commands you're going to use when

configuring the classes for queuing it's

either priority or bandwidth priority is

the command you use when you're doing

priority queuing bandwidth is used to

allocate bandwidth to classes that are

not priority classes you know this is

our class-based weighted Fair queuing so

this is going to be a priority queue so

I'm going to say priority question mark

now I can either specify a percent of

total bandwidth

or a value in kilobits per second

Cisco is going to tell you that you

should not have in most cases more than

30 percent of your bandwidth

allocated to Priority cues so we're

going to keep this number low I'm going

to say 10 priority 10.

I would have been diving 10 kilobits per

second which is not what it was so

priority percent 10.

um guess what that's it you know we've

already applied this the service policy

um to you know in theory our egress

interface we've now said priority

percent 10 you know there you go if I

want to look at the llq configuration

we can do a show policy map interface in

fact let me look here to see which

interface I applied it to I want to say

it was one of the ethernet ports

it was

oh

oh one okay it was o1 so if I do a show

policy map

interface fa01 you're going to see what

we've configured now

if you look at the class map of wave RTP

match any

we can show the matches that are being

made we see metrics on the matches being

made how many packets have matched

and we see

that we are doing priority

10 percent

pretty cool huh

and it's based on the interface speed

obviously so right there we've

configured the priority queuing part of

llq

now for other classes

I want to configure class base weighted

fair queuing so let me let me do a show

Ron on the uh

section policy map again so we've done

the class VoIP RTP and the other class

let's look at the

class map as well because we'll need to

reference that so we'll go conflicty

we'll go to policy map qos and we'll say

class

VoIP Dash control so this is our

signaling traffic

um

we can then use the command bandwidth

question mark and we can either say the

amount of bandwidth in kilobits per

second

we can use percent of total bandwidth

or we can use the you know remaining

bandwidth

um

I'm not going to tell you exactly what

values to use here because this is going

to vary from Network to network in fact

you need to do some calculations

um

relative to the specifics of your

network to figure out how much bandwidth

you want to allocate so we're going to

just kind of keep that simple here I'm

going to in this example allocate that

as some percent five

so

if we do a show

run pipe section policymath you'll see

for each class

what we're doing so for the class Vape

RTP we're setting the dscp tag it's EF

perhaps it already wasn't perhaps it was

and that's unnecessary but you know

priority percent 10 so that's priority

queued

class VoIP control

bandwidth percent five so that's going

to be class based weighted Fair queued

pretty simple right now if I wanted to

Define other classes I would simply

create those classes

and use the bandwidth commands

you know or the priority commands up to

30 of the interface bandwidth but uh use

the bandwidth commands to allocate

bandwidth amounts to my individual

classes that I've defined so potentially

there's you know eight classes in my

model

probably one or two of them are priority

queues everything else is class based

weighted Fair queuing divvy up the

bandwidth however you'd like to divvy up

to the bandwidth now when you use the

bandwidth command

um

it's going to

um let's see what's the best way to

describe this

if congestion occurs because remember

this stuff's only happening when

congestion is occurring if congestion

occurs

traffic exceeding the specified

bandwidth is going to be dropped

um

yeah that's probably the best way to put

it

if it's not

congested

then you can continue to use bandwidth

but if it's congested

traffic in excess of the configured

bandwidth value will be dropped

it's going to be tail drop by default

but if you're using something like

weighted random early detection then

obviously you can specify drop

characteristics

pretty straightforward there so if you

want to I already actually showed you

the show coming from honor and let me

run it again and show you the

policy map is applied and let you see

you know at first we had this VoIP RTP

prior to 10 and now we've got the VoIP

control

and we've got

bandwidth five percent so five thousand

KB

and then we've got the class class

default that's going to be everything

else

so pretty straightforward

um you just continue to build on this to

make a larger model in fact we'll show

you an example of what one configuration

could be for that qos Baseline reference

we talked about that 11 class model

again your network is going to dictate

the exact values but this should be a

nice you know a nice um you know high

level overview of what's going on with

the basic llq and then class based

weighted fair queuing is participating

in there

now that we've covered the queuing

concepts we're going to talk about

traffic shaping and policing as we go on

so thanks for watching I'll see you in

the next video and good luck with your

studying

[Music]

I originally intended to cover traffic

shaping and policing as part of the same

video where I discussed queuing but

things just got a little bit long and I

didn't want to drag it out and do a big

30 minute clip so I've decided to cut

that to that first video short and we're

going to pick it up here with the

shaping and policing part of the theory

and configuration on the router so with

that into traffic shaping and policing

yeah

let's talk about traffic shaping and

policing and the first question that I'm

going to present to you and the first

thing I'm going to try to explain is

first off what's the difference

queuing only gives us so much you know

it gives us the ability to

order

or schedule how packets are placed into

a queue and placed on an interface

but it doesn't really allow us

to

deal with traffic

in excess of of what we're trying to to

you know transmit

or manage

traffic shaping and policing give us the

ability to manipulate the data flows in

a way that queuing alone cannot so first

off when we talk about policing

a policer is typically going to be

involved in discarding or dropping

traffic whereas a shaper is typically

going to be buffering excess traffic to

transmit at a later point in time so as

we talk about policers and Shapers you

need to understand how this evaluation

takes place and you know the exercise we

go through

shaping and policing can work together

as part of your Enterprise qos policy to

manage traffic

in the way it's most appropriate for

your business and for your

infrastructure

you need to understand some basic

concepts about traffic rate measurements

and the methods used to measure traffic

rate

you know which is going to be

at the Forefront of any policing or

shaping activities

first off the concept of a token a token

gives permission to send X number of

bits a token bucket can hold a specific

number of tokens and really the

relationship here is that a token bucket

can hold so much data

so let's start with that tokens arriving

after the bucket is full

are discarded and are not available to

Future packets so once you fill the

bucket the bucket's full any more data

you throw at it is gone

if there are not enough tokens in the

token bucket to send the traffic I.E if

the bucket is full and the traffic

you're trying to send exceeds the amount

of tokens available in the token bucket

you can either discard the packet and we

call that policing or you can continue

to wait for enough tokens to accumulate

to give you the ability to send the

packet and we call that shaping

there are some reference indicators we

need to talk about relative to traffic

rate there is the BC or what we call the

committed burst rate

and that's the amount of data guaranteed

to be delivered within a given interval

there is a be which is an exceed burst

or an excess burst and that's an

additional amount of data that you'll

permit to be attempted for transmission

if there's no congestion

the TC is the time interval and you've

heard the term committed information

rate and that's really what shaping and

policing about is matching up to a

committed information rate and a

committed information rate can be

calculated as the committed burst rate

divided by the time interval

a bucket size the token bucket in total

can be conceptualized as the committed

burst rate

plus the exceeded burst rate and we're

going to get into this whole token

bucket model and I'm going to sketch it

up and explain it to you but let's let's

make sure we understand these these

basic concepts

typically under normal use at a normal

rate

the committed burst rate so much data

whatever that rate is is sent every time

interval

on exception or some of the times or I

should say when permitted

the committed burst rate plus the exceed

burst rate can be transmitted if enough

tokens have been previously accumulated

in the token bucket to understand how

policing and shaping work you need to

understand the token bucket model and

token buckets you know and really play

into

how decisions are made you know we need

to figure out am I meeting the committed

information rate or am I not meaning the

committed information rate and you know

if I'm not meeting it or if I'm

exceeding it what do I do so let's start

by talking token buckets so a token

bucket quite simply we'll see if we can

draw a

that's supposed to be a bucket so that's

my token bucket and it has

a certain capacity

of tokens let's say it is 1 000 tokens

and say that it's actually a thousand

bytes

worth of tokens

if a packet comes in

that is one thousand bytes

and I have a thousand tokens

then I'm good you know I conform

to

the token bucket Raider right I'm

conform to the committed information

rate and the packet is forwarded

so if I don't

then I don't have enough tokens and the

uh you know the packet will not be

forwarded

the

way we scale this into policing and

shaping

is by

concerning ourselves with the rate at

which packets or tokens I should say are

placed into the bucket and passed

through the bucket so in a single

token bucket model

the token arrival rate or the cir the

committed information rate

is going to flow into the bucket so I

showed you here that was a thousand

bytes in this example right there

thousand bytes

and this this value the depth

is our

burst commit or our commit adverse rate

so

the fact that we have

a

input rate that conforms to our

committed burst rate

means our traffic will conform now let's

say it doesn't conform let's say I have

more than a thousand bytes let's say I

have you know 9 000 bytes what do I do

well

if I can form great if I don't conform

I'll say no if I don't conform to the

rate

then I can throw that packet

into a second bucket

and

I can configure

the burst rate

or the exceed rate let's say exceed

of that bucket

if

I am

small enough

to be underneath this maximum exceed

rate

then I exceed

now that's not a bad thing

you know that's that's kind of like a

success

if I am not and I'm still too big

then I violate

and that's bad violating is bad

so not the prettiest sketch hopefully

this makes sense what's going to happen

with

these methods can form

exceed

and violate

is we're going to be able to configure

our qos Behavior or our shaping and our

policing Behavior

based on whether or not we conform we

exceed or we violate so conforming just

to sum it up means there are enough

tokens

to

you know in the first token bucket that

we've sent it to

to conform with the maximum size of the

BC

exceed

means there are not enough

in the first bucket

however

there are enough

in the second bucket

a little slow writing tonight

that we are under I should write it in

here DB e the burst exceed

if we are violating

it means there are not enough

in one or two okay so token buckets very

important as we jump into the next

section here or the next slide we're

going to go into the router

and talk about configuring policing and

shaping and go through some examples and

show you how to leverage these

conform exceed and violate actions

all right so we're back in the router

and we're going to configure

some shaping and some policing now let's

do policing first when you know when you

think about policing we're going to

limit the input or output transmission

rate

relative to a particular class of

traffic that we have defined

we can use either a single or double

token bucket method

and

you know so that we can specify you know

violate actions

if you use a dual token bucket method

you're going to be able to conform to

the rate limit when the traffic is

within the average bitrate you're going

to be able to exceed the rate limit when

it exceeds the average bit rate but

doesn't exceed the excess burst rate and

we're going to violate the rate limit

when the traffic exceeds both the

average and the excess rates so just

like the sketch that I showed you

and basically we're we're um we're

marking things we're making decisions

based on what happens

could I choose to transmit traffic

all the time

that exceeds my cir

from a router perspective yeah I could

do that my service provider might have a

problem with it or my uh my billing

department might have a problem with it

when they see the bill of um you know

that when when the bill comes in because

I've been bursting above my cir that I'm

paying for

so let's let's get into a really quick

really quick policy map here so let's

actually let's define a class map and 50

we're going to say class map and we'll

just call it

um

you know my traffic and we'll just say

match protocol HTTP that's that's good

enough now we're going to go policy map

and we'll just call it qos and we'll be

in that same policy map and we'll say

class

my traffic now I'm going to say let's

let's do policing first police and I'm

going to say either bits per second

the cir or the rate I'm going to go

ahead and do this one in bits per second

206.00 so there's my bits per second now

check this out I've got conform action

so I can specify what to do when I

conform

so I can say

transmit I can you know manipulate the

SCP values and then transmit so maybe I

could you know if I can form I'm

probably not going to mess with these

cpux but if I exceeded you know maybe

I'll lower the qos level down

so but for now we'll just say transmit

if I exceed action

maybe I transmit anyway okay fine maybe

I drop

maybe I remark the packet or remark the

dscp value

so that Downstream this is treated

differently

if I'm going single token bucket

um you know we've got the conform and

exceed we will say drop

so there's single token bucket so show

run

we'll just go full blown Show run and

I'll show you what's important here

so we've got our normal

class map you know we've defined a new

class here we called it my traffic

and we said match protocol http

we're using the same qos policy map that

we've been using but for the new class

we've defined we set up police at a rate

we said conform action transmit exceed

action drop

that's single token bucket policing

if I wanted to use a dual token bucket

model we'll go back into that policy map

and this time

oops hang on class my traffic

and then we will say exceed action

transmit

violate action drop

so now show run

get down to it here

we will have a conform

and exceed and a violate now you've got

the ability

to

um

yeah really it's kind of up to you as

far as how you want to do this and what

you want to do

but very very cool

I wish I could give you some best

practices on policing and uh and shaping

but it's really so unique

to each environment

you know I I had one time you know an

example that comes to mind is I had a

Wan link and I had a really really

chatty database connection

and it was just hammering the snot out

of my landlink and I I needed to cap it

I did some shaping and policing to keep

that traffic under control and I

actually did some you know a lot of

dropping and uh it solved my immediate

problem but uh certainly started some

other conversations

up you know that's basic policing

if I wanted to get into shaping

and try to manipulate things a little

bit more I'll go ahead and do it for the

same class of my traffic we'll go config

t

and we'll say policymap qos class my

traffic and then we'll say no police

actually no let's get rid of all this

boom

gone okay we will this time say shape

and I can do shaped average

or shape to Peak I'm going to go ahead

and shape to average

I again Define a Target bitrate

or a percent of the interface bandwidth

for the configured committed information

rate but so we'll go ahead and we'll

we'll be explicit here

and let's see what's a good rate to use

let's say 32 oh oh

and now we're going to shape it see how

easy that is if I wanted to do a shaped

Peak shape Peak again you know 32 oh oh

then I can do a shape Peak and

what's interesting about shaping versus

policing

is the goal in fact let me kick back

over to the screen here the goal of

shaping if you look at a graph here

you've got your traffic

and you've got

um you know policing Behavior so this is

our traffic rate

and I'm going to show you shaping

the goal of shaping

is when the traffic levels are

increasing is to try to level them off

if I'm doing policing

again the same chart

what's going to happen is as traffic

ramps up

it's going to drop

like this

so that's policing so I'm gonna

cut it off when it hits a high enough

level whereas with shaping I'm going to

try to massage it and smooth it out

um

policing you know when you're dropping

these packets obviously you're going to

be causing you know it's data not

getting where it needs to get so you're

going to have TCP retransmissions

happening

um because you're shaping you know in

the other example

you're going to minimize the number of

TCP retransmits so you know there's

there's places for both

and you know again I wish I could give

you some more you know best practice you

use but it really is business case

dependent so those are the foundation

Concepts that Cisco is going to want you

to understand

relative to traffic shaping and policing

don't make it harder than it has to be

it's really a lot more work to explain

it

than it is to configure it you know like

I showed you

um you know a couple lines of code here

is really all we had to do in fact let

me do just for another zoom in on things

here a show run

and let's get down to our policy stuff

and you know right there

there's our class of traffic and there's

our shaping configuration so that's

pretty much it um I think that's going

to wrap up this section of the video

we've got one more to do and that's

going to be talking about Cisco Auto qos

and it sounds exciting it sounds like

gee Josh if we have this thing called

Auto qos and I can just turn it on why

do I need to know all this other stuff

and we'll talk about that we'll answer

that very question and should be a good

um you know a good way to wrap up and

round out your Cisco qos toolbox

knowledge and prep for this C voice exam

and then finally we'll have a video and

we'll talk about an example of how to do

all this stuff we've done combine

relative to that large 11 class Baseline

qos model that we demonstrated earlier

so with that I'm going to say thanks for

watching I know we've covered a lot of

information here this is definitely one

of those videos you're going to want to

rewind

and uh go back through I've tried to

um purposefully slow down my description

I know I tend to go a little fast and

this is one of those that you kind of

need to creep into so I've tried to talk

a little slower and not rush myself but

uh feel free you're not going to hurt my

feelings if you press that the rewind

button and play it over so I'll see you

in the next video and good luck with

your studying

[Music]

welcome to module 38. in this module

we're going to be talking about Cisco

Auto qos and I want to tell you that

we've reached or you have reached pretty

much the last instructional model within

the course sure we're going to have one

more video following this where we talk

about

um you know the 10 class I'm sorry the

11 class Cisco Baseline qos model and we

go through a complete review of the

configuration of the model but really

that's a bonus you know you've seen how

that works already so that's not

building new skill you know you've got

that skill already

Auto qos is kind of uh you know wrapping

things up for you now let's talk about

Auto Qs you know we went through the

modular qls CLI and mqc and took the

time to understand

um how we can Implement queuing methods

using class-based weighted fair queuing

and low latency queuing we took the time

to understand

how layer 2 and layer 3

qos mappings worked and we went through

examples of shaping and policing and we

really covered you know how class maps

and policy Maps

work together to provision an Enterprise

qos deployment so what is auto qls and

why are you telling me about or why am I

telling you about it now at the end of

that well the answer is simple because

if I showed yado qos you'd never take

the time to learn what's really

happening behind the scenes and you need

to you know understanding qos and

designing Enterprise qos strategies is

important it is something you're going

to do quite a bit of and you really need

to understand it well so what is auto

qos and why am I going to use it instead

of taking the time to design those

Enterprise qos strategies that I just

talked about well the reality is in the

industry you're going to run into a lot

of different networks and they're all

going to have different characteristics

and you or the client may not have

adequate times Staffing Resources

whatever to

fully understand their network but yet

they still need to have a a quality Qs

deployment done

so what do you do auto qos is kind of

the next best thing to and dare I say it

this way but to uh to doing it right

that's probably a bit unfair for auto

qos it's you know Auto qos is doing it

right it's just not doing it manually so

let's talk about what is auto qos so

we've got two components to Cisco Auto

qos there's Auto qos VoIP and auto qos

Enterprise and auto qos VoIP is a

technique in fact it's you know it's the

earliest thing that we called Auto qos

and it lets you quickly and concisely

deploy qos policies to your routers and

switches

with configuration applicable to voice

transport

it really doesn't take into effect the

rest of the Enterprise it's really just

a kind of a voice thing

Auto qos Enterprise you know takes that

concept to another level and helps you

automate the deployment of qos policies

using a 10 class model so instead of

being just kind of watered down for

voice auto qos Enterprise you know helps

you step it up a little bit and auto qos

voice or VoIP is you know a couple of

commands to deploy and it's going to

create class maps and policy maps and

apply those service policies to your

interface and do those things for you

Auto qos Enterprise is a little bit

different and that it's going to

actually go through two phases you're

going to take it through an auto

Discovery phase where it watches the

traffic on your network and figures out

what you have going on and then there's

a provisioning phase where it actually

creates these policies and structures

things out so we'll take a look at the

auto qos configuration Auto Qs VoIP is

where you need to focus most of your

effort or most of your time on this

but know about Auto qos Enterprise and

that it exists now I'm going to walk you

through some configuration of Auto qls

and let me pull the pull the our good

old HQ router here the 2811 and let's

walk through some basic audio qos

configuration

if I do a show run if I can type it

right here Show run there we go

I'm going to show you that there are no

class Maps

and no policy maps created on this

router it's blank from a qos standpoint

for me to enable auto qos VoIP

on a router what I'll do is I'll go to

the interface

of Interest let's let's act as though

this is my Wan interface that interface

of congestion so we'll go to config

interface0 colon zero and I'm going to

say Auto qls

VoIP and then I have a choice to make I

can either trust the dscp markings as

they come in on the interface

or

I can

not trust it and I'll use nbar to detect

the application

and kind of do my own thing I'm going to

not trust I'm just going to go ahead and

let anbar do its thing nbar support is

required for auto qos and for nbar to

work you must have CEF enable Cisco

Express forwarding so keep that in mind

this is supported on most of the gear

you're going to run into and actually

following command was not properly

applied service policy output on a qos

policy untrust I have no idea why it

couldn't apply that let's take a look

and see if we can find out why

let's see let me show you what it bill

serial zero zero

zero zero

I actually don't see

any real reason why that would have

failed let me get rid of we're just some

garbage here this has nothing to do with

the failure but let me get rid of this

no map class frame relay frame just to

get the garbage config out of our way

let me go ahead and put an IP address on

that interface interface zero zero zero

zero colon zero IP address uh wants to

know two sixteen Five Dot one two five

five two five two five five two five two

no shut let's see if it lets us do it

now it's possible that because I didn't

have an IP address on the interface that

it didn't want to run auto qos VoIP plus

cross our fingers and see if this works

I'm hoping that it does

it's grinding along here

no it still doesn't service policy

output auto qos policy on trust that's

weird you know let me do this

um let me go ahead

and try this on an Ethernet interface

just to see what happens and FAO One

Auto qos flape

and really you know the command I'm

showing you that's how you enable you

just turn it on

and it's going to generate the class

maps and policy maps and things like

that for you now that worked let's take

a look at the show run I'm going to show

you exactly what it generated and what

it built and then we'll continue to dig

into what that error is coming from and

why it's causing it so if we look at our

configuration Auto qos has generated

three class Maps a class map match any

auto qos VoIP remark

and you'll see that it's matching

traffic with VoIP related dscp tags so

efcs3 and af31

we have a class map match any for auto

qos VoIP control untrust

which we're going to be matching in ACL

and we'll take a look at that ACL here

in a second

and that's called audio s flip control

and then we've got a class map match any

audio qos VoIP RTP untrust

and then we're

going to protocol our TP audio using n

Bar and then match an ACL so okay so

we're looking for stuff that hasn't yet

been trusted

and trying to classify it with nbar down

there and we're trying to classify with

an ACL presumably based on port numbers

and stuff here so pretty straightforward

so we've created some class Maps

we've created a policy map called auto

qos policy untrust and we've got class

Auto qoswipe RTP untrust priority

percent 70 set dscpef so oh boy we're

doing a lot of priority queuing

allocating a lot of bandwidth on this on

this interface for RTP audio so that's

definitely a voice-centric thing

Wave Control bandwidth percent five so

we're doing some class-based weighted

fair queuing and we're setting the dscp

tag in fact we're doing that up here in

the RTP as well

and then we've got the VoIP remark set

dscp default so we're basically in this

one if we see these tags we're going to

go ahead and clear them and set the dscp

marking back to default so interesting

class maps and policy Maps let's keep

going down and let's look at our

interfaces oh I bet I know why it wasn't

working it's because it's a ppv

multi-link and I would actually want to

do it on the multi-link interface not on

the individual stereo so let me let me

do this real quick let me back out what

I did I'm going to go to interface fao1

and we're going to turn off auto qos

points just simply by saying no autocue

swipe it's going to rip out those policy

maps and class maps that it created and

we're going to go to this interface

serial zero zero zero colon zero and

we're going to say no PPP multi-link

so that should be out of there now

let's see if it is yes it is that's a

normal link now enter zero zero zero

zero colon zero Auto qos void enter

Let's cross our fingers that should

apply successfully now that has to have

been why it was erroring out I can't

think of any other good reasons

ah let's see here ah fantastic okay so

it was just a problem with the fact that

I had PVP multi-link turned on so good

troubleshooting exercise there so show

run again nothing's changed as far as

what class maps are going to be created

so we've got the same class maps that I

described to you and we've got the

policy map created so again it's a

priority queuing for llq we've got some

class-based weighted Fair queuing and if

you look at the interface that serial

zero

zero you'll see auto qos flape is there

and it's attached the service policy

output Auto qos policy untrust so it did

a traditional modular qos CLI style

configuration leveraging the parameters

that we configured so we weren't

trusting the dscp data that the router

had we were configuring

marking you know based on our own

protocol detection using nbar

so definitely pretty cool

um there's nothing else that it really

built you know there's these ACLS I was

talking about right here it built those

and like I said before the VoIP control

we're doing some preventive TCP and UDP

so you can see we've got some key ports

in there you know 50 60. you know we've

got some you know for sip we've got some

skinny

um yeah I don't know what all these

ports are without looking them up but uh

VoIP control related ports and then

we've got some rtcp stuff here as well

so really Auto qos Vape on a router is

nothing more than automating the

deployment

for you it's not any real magic it's not

doing things drastically different than

we do if we did them by hand it's just

not giving the level of granular

attention that we would give to a manual

configuration so

is there anything wrong with using Auto

Cree OS absolutely not if you're not

going to turn on qos then at least turn

on auto Qs audio Qs is great for the guy

who has 200 sites and one it guy

and that one it guy is never going to

become a qos expert but he knows he

needs it and he needs it to be

consistent across all 200 sites so he

turns on auto qos the router takes care

of the rest and life is good so that's

Auto qos vape and that's really all you

need to understand about it to configure

it on a switch in fact standby one we

will open another putty window here into

the 3550 switch that we've been using

and we'll drag it down here to about the

same size

a little bit bigger

and to configure Auto qos on the 3550

switch I'm just going to say Auto qos

am I even supporting on this version of

iOS you know I may not be

let's see

I actually have to turn on qos globally

first so MLS qos

can I just do that and then go to the

interface we'll say interface fao1 Auto

qos

VoIP yes I can and then we'll say Cisco

phone

so what I've done here

is on that interface fa01 Show run at

fa01 I have

created

all of this configuration relative to

Auto qos so you can see I'm doing some

weighted round Ramen queuing uh class of

service Maps we're doing some bandwidth

allocations for waiting around Robin

queuing we're actually trusting the dacp

tags if we detect using CDP that the

connected device is a Cisco IP phone

um so you know pretty straightforward

stuff obviously I could do this to you

know all the access ports and you know

I'd want to you know do some things a

little different to the Uplink ports uh

relative to trust or not you know

depending what is called for in my

environment but Auto qos is there you

know it's that easy to configure there's

not a whole lot else going on in the

switch let's take a look you know you

see that we've got the uh the basic MLS

qos configuration we've got our cost of

the SCP mappings set up

um

you know we've got that one port we've

configured but other than that you know

things are pretty much vanilla relative

to the configuration of this device so

very very cool stuff

and uh really that's all I want to show

you about Auto qos VoIP

um let's go over real quick

um to the uh actually one more thing I

want to show you

before I leave the switch show Auto qos

enter you can see fa01 auto q s voice

Cisco phone so it's showing you the

interface that I've got uh turned on for

auto qos so cool stuff

now if I want to do auto qos Enterprise

let me get out of the switch I'm done in

there and back to the router for auto Qs

Enterprise things are a little bit

different in fact let me back out

interface zero zero zero zero colon zero

no Auto qos get rid of Auto qos swipe

and we'll walk through remember I said

not a Qs Enterprise Has Two Faces you've

got Auto Discovery and you've got

provisioning to get into Auto discovery

mode we're going to go to the interface

actually I'll go ahead and use that same

serial interface and we'll say Auto

discovery qos

and then again either trust or don't

trust the markings so we're gonna not

trust the marketing so we'll use n Bar

to decide what's going on

and typically you're going to want to

run this discovery on the ports of

interest uh for a couple of days you

know

two days three days a week you know

through the course of your normal

business cycle

and uh things are going on here all

right cool so Auto Discovery qos has

been turned on once we're done

um then we're going to need to accept

the configuration after we've done the

discovery so we're going to do that by

saying Auto qos

enter now I haven't had any data so it

didn't discover anything let me try

another interface

just to uh

to make this a little more interesting F

interface let me do the show IPM brief

um

how about

we do it on f8

00.30

we will say Auto discovery

is it gonna let me run it on the sub

interface it may not I may have to do it

just on the fa 0 0.

Auto Discovery yeah I'll have to do it

there so interesting can't do it on a

sub interface I did not realize that so

we've got AutoCAD Auto Discovery qos

running

and we'll just let that run here for a

few minutes

and that's it let me see if it lets me

uh Go Auto qos yeah no Auto Qs data

discovered I'm not letting it run long

enough if you were going to use this in

the Enterprise like I said before you're

going to let it run for you know two

five seven days whatever

and uh what's gonna happen is once I

generate the policy based on what I

discover that auto qos command that I'm

typing is going to commit it to the

device and it's going to actually have

built multiple classes for us in the

policy and there's some commands it's

probably not going to give me a lot of

output at this point since we haven't

discovered anything but if we do

show Auto discovery qos

it actually does let us let us see a

little bit here so you can see for fast

e00 and the serial you know the

discovery up time in fact if I refresh

that you'll see it's continuing to clot

I just don't have enough traffic running

through this router to generate anything

but what it's going to show you is what

classes have been discovered and what

it's creating it's going to show you

recommendations for minimum bandwidth

detected applications

um

really just kind of break it down for

you and you know it's going to build you

um

like I said before a 10 Class A qos

model not all that different from the

Baseline qos model that we've talked

about Cisco using so but you don't need

to get a whole lot deeper into this for

the C voice exam I want you to know it

exists they want you to know how to kick

off a discovery and commit the changes

beyond that you should be should be good

to go so as we wrap up this final video

um you know and again I told you if

there'll be one more I'm going to talk

about that 11 class model and show you a

working example but as we wrap up this

video we've wrapped up the the heart and

soul if you will of what you're going to

need to be able to prepare for the C

voice exam like I always say one

resource whether it be a class that

you've taken a book that you've read or

a video series that you've gone through

no one source is going to adequately

prepare you for this you know any

certification exam uses many sources of

material as you can use on-the-job

experience use your own lab or rack

rental or you know take a class and read

a book and watch a video of course all

of these things are going to help you

know well-round you off as an engineer

and prepare you to leverage these

Technologies in the real world and

ultimately pass the certification exam

so

um I really appreciate you hanging in

there with me I know that we've covered

a lot of material you know sometimes

it's hard to demonstrate these things

outside of the real world environment

but we've tried to do that and that's

pretty much it for the the C voice exam

course so uh hang in there take a look

at my Enterprise uh

Baseline qos video next and uh after

that the review review rewind watch the

videos as necessary and good luck on

your exams you know as you go into the

test I want you to think about it as an

opportunity to

um

prove to yourself what you know

I am not the best exam Taker and it

takes me on sometimes several attempts

to pass tests and there's nothing wrong

with that you know create a study plan

that works for you if you don't pass

right away the best advice I can give

you is to go back study immediately fill

in the gaps and go take it again do not

wait two three four weeks do not wait a

month don't do that take it right away

and you're going to have the best

preparation to pass on your next attempt

so thanks for watching guys and I'll see

you in the next video and in the next

course

[Music]

thank you

[Music]

you've made it you have reached the

final video

we're going to talk about the Cisco

Baseline qos reference one more time and

I'm gonna give you a configuration

example I know we've gone through a lot

of different qos stuff and we've shown

you examples of shaping and policing and

queuing and building class maps and

policy maps and we've shown you Auto qos

VoIP and Enterprise and I want to round

It Out by showing you how I do things

this 11 class model I certainly did not

invent but I've adapted to my own

purposes throughout the years and I've

had really really good luck with it I

have to credit

um Tim segetti one of the ccies that I

know for

um including this design in a book he

wrote several years ago end-to-end qos

Network design in fact I think he's got

a new addition now I need to pick it up

but that book really went a long way at

explaining qos models to me

and he thought enough of the model The

Cisco reference model to include it in

his book and I have adapted it slightly

and included it here in the video so if

we look at the 11 class model I've got a

breakdown of the types of traffic we're

going to be including we're accounting

for voice video conferencing streaming

video call signaling routing Network

management traffic Mission critical data

transactional data bulk data scavenger

traffic and then best effort we are

going to allocate bandwidth based on the

percentages shown in this graph and keep

in mind these values are not set in

stone these values can vary from

Enterprise to Enterprise and and should

they should be set up to match your

unique environment that said this model

can give you a solid starting point for

doing your own design looking here the

things I want to point out

is we're using voice video conferencing

and streaming video are all going to be

priority queued so 18 plus 10 is 28

percent

and uh five percent for streaming video

is 33 and I mentioned earlier that you

want to stay

um not to exceed about 30 the actual

number is 33 but uh as far as what

you're going to strict priority queue or

loq but um so we've done that here and

really I'm gonna walk straight into it

we're going to show you the class maps

and the policy maps that you can use to

implement this 11 class model so we've

created class maps for each of our

classes we have class map match any

network management match IP dscp CS2

class map match any bulk data match IP

dscp af11 and af12. class map match any

voice match IP dscp e f class map match

any Mission critical data match IP dscp

25.

class map match any round team match

ipdscp CS6 class map match any scavenger

match IP dscp cs1 class map match any

video conferencing match ipdscp AF 41

and 42. class map match any streaming

video match ipdscp CS4 class map match

any call signaling match ipdscp CS3 and

af31 and we've got a class map match any

transactional data match ipdscp af21 and

af22 again this is not an absolute you

can adapt this model as needed for your

environment I'm not using in this

example anyway I'm not using nbar I'm

assuming that my packets coming in are

appropriately marked now that may or may

not be the case in your environment so

keep that in mind as you go through your

configuration examples

we've got a policy map that matches up

to these class Maps we're going to call

it policymap qos and we've defined for

each of the classes in entry so class

voice priority percent 18 and remember

I'm using the priority command that

means llq or low latent security

class video conferencing priority

percent 10 class streaming video

priority percent five

now we get into the class-based weighted

Fair queuing where we've got the

streaming video class with five percent

the call signaling class with five

percent routing with three percent

Network management with two percent

Mission critical data 15 percent

and keep in mind think about that

if this were up to your business

without technical input how many percent

do you think they'd allocate to Mission

critical data I guarantee you it's going

to be way up there it's not going to be

15 they look at that and they'll say

what are you nuts but really this is

appropriate you need to play that

liaison and and uh help guide them

towards a good solution so we're showing

you here 15 and we're doing random

detect

class transactional data bandwidth

percent 12 with random detect dscp based

bulk data has four percent again random

detected dscp based in our scavenger

class bandwidth percent one so we're

really dropping that down

and then class class default that's the

everything else bandwidth percent to 25

random detect so I think this is going

to give you a good solid model to build

on

this is a fantastic model for a branch

router or a core router doing Wan

services and I think you can build on

this for your own qos policy as

applicable to your network and when I

mentioned

partially deploying the 11 class model

you may not build every one of these

classes

but you should you know consider the

possibility of leveraging the same

values anyway just because you've got a

hundred percent

if you only choose to implement class

voice and class call signaling

you're you've accounted for 18 plus five

percent you got a lot of percents left

I wouldn't recommend changing the

numbers

leave yourself That Elbow Room so that

later you can come back

and add additional classes as you choose

to leverage them

in your environment so this is a good

example to start with it's not you know

it's not something I would say copy

paste and run within your router in fact

I'm not even going to do that here in

the lab you know it's a reference use it

as one and hopefully it'll help you have

a good gut feeling of where to begin so

this is it this was kind of a little

bonus thing

um thanks for tuning in to this video

series I think we've had a lot of fun

demonstrating how things work sometimes

we've had a lot of fun demonstrating how

things may not always work and

troubleshooting techniques and gee why

isn't that working you know and and I'm

sitting here and I'm I'm recording and

I'm screen capturing and it didn't work

perfect the first time so what do we do

we're not going to stop the video we're

going to let it play we're gonna go

through the troubleshooting thought

process and the exercises and we'll

figure it out in fact we had a pretty

good one about a video or two ago where

we were working on the serial link so

that stuff all adds value to the

instruction process it's the kind of

stuff you're going to run into in the

real world and you need to understand

how to deal with it so I like to leave

it in there when possible when I'm not

looking like a complete and total

so and you know even if I am that's okay

but uh good luck with your studying

process I hope that this video Series

has been helpful to you we will be uh

creating video series for the rest of

the Cisco ccnp voice track in fact next

up is cipt1 which is primarily a call

manager class so uh hopefully by the

time you have passed your C voice

content well or your C voice test will

have the cipt1 content ready to go and

you'll be able to jump right on in and

then keep moving with your certification

process my name is Josh Kittle it's been

a pleasure being your instructor for

this video series and I hope to see you

in the next video series good studying

and thanks guys