All language subtitles for 2. Another reason to use a loopback - OSPF

A previous blog I explained why you would use a loop interface and I gave you one reason.

Let's look at another reason.

Another reason why you want to use a look back is because writing protocols such as always use the loopback

to determine the ideas of routers in the network.

So these routers

show IP interface.

I've configured a loopback zero with this IP address and I've also configured the loop banks on each

router Sarada one is 1 and 2 1 6 8 1 2 1 router Tuesday or 2 rotisseries or treat.

Now why is that important.

It's important for routing protocols.

So when we enable writing protocols such as SPF it selects a router ID for itself.

So I'm going to enable OSPF on all interfaces on the router and just simply put them in areas Zira.

So show IP SPF interface brief SPF is enabled on all interfaces on the router show IP SPF interface.

Notice show loopback is 0.

Has this IP address configured in area 0 but the router ID is 1 9 2 1 6 8 1 to 2.

In other words the Rodda ID selected is the highest IP address of any interface.

And if they are loop backs the loopback override the physical interfaces.

So the Rodda IDs are elected on the highest loopback address.

In this case 1 9 2 1 6 8 1 to 2.

Now that's important because in OSPF a router is identified by its Rodda ID it's kind of like your name.

So my name is David.

It would be terrible if my name changed everyday whenever there was a problem.

And that's the issue with Louise.

If you don't use a loopback interface and the Rotto I.D. was selected off a physical interface and that

interface when down the road the idea would change but it's consistent.

Notice on gigabit 0 1 the router ID is still 1 2 1 6 8 1 2 2.

So let's have a look at the problem.

Not a one.

We've got two IP addresses configured on physical interfaces and then we've got these loopback interfaces.

But what I'm going to do now is remove the loop back into faces of Route 1.

So show IP interface brief do that again.

We only have IP addresses on the physical interfaces.

So now Rodda OSPF one network and I'll just enable all interfaces once again notice a neighbor relationship

has established.

We've actually got two.

But notice the neighbor ID is 1 2 1 6 8 1 to 2 ways on this side the neighbor I.D. is 10 that one to

two to one the highest IP address of a physical interface.

This loopback doesn't exist on route one so on router to show IP who is P.F. neighbor hibernation ship

is full and established tendered one to two to one right one show IP OSPF neighbor notice a difference.

Neighbor ID is 1 and 2 1 6 8 1 2 IP address on the interface that the neighbor is using it is tendered

one or two to two.

And on this interface gigabit 00 extended one or two.

So the same neighbor ID the loopback.

And just to make sure that its not confusing at that here as well weve got the loopback configured on

rodded too.

In addition to to the to the to the ATI brought it to now when we look at the neighbor relationship

with one neighbor ideas tend toward one or two to one that's this interface.

Gigabit is one the IP addresses and the physical interfaces are those on road a one.

So why is the wrong idea important.

Well when an interface goes down and the road is reputed the Rada ID will change and in certain cases

in a whisper if we specify the Rodda ID in commands one of those cases which you don't need to know

if DNA is a virtual lynx in this command.

We are specifying the router ID of the remote router a virtual link allows us to create a virtual transit

across an area.

In this example of backbone areas on the left area one is in the middle an area two is on the right

area.

Two is separated from the backbone area by area one but a virtual link allows us to tunnel areas zero

across area one and full one of the requirements of always piaffe which states that all areas have to

be adjacent to areas zero.

Now the problem is you know typology at the moment and wrote a one show IP OSPF interface.

Notice the right of ID is tendered one to two to one.

So that's the route ID of the router and you can see that on both interfaces what I'm going to do however

is shut.

The zero 0 1 down so currently show IP OSPF interface shows that this is the route ID if the router

rebooted the Rodda ID would change instead of rebooting the router.

I'm going to type clear IP CPF process and I'm going to clear all processes

so show IP OSPF interface.

In this case the right ID didn't change.

So what I'll do is save the configuration and reboot the Rodda.

So why is that a problem.

If the interface that you're using was selected by the Rodda at the Rodda ID and you configured that

draw to ID and this command and the interface went down and the rotor rebooted this command would no

longer work because the Rodda Id had changed.

Noticed Cisco Telli the router Id usually the highest IP address on the box or the highest loopback

address if one exists.

The run ID is only calculated at boot time at any time when the process is restarted.

So what's the recommendation.

For OSPF specify a loopback address and then manually configure the Rodda ID.

So the Rondo's reboot show IP OSPF interface.

Notice the router ID has changed.

It's now 10.0 wondered wondered what that means if you're using a virtual link as an example the virtual

link would break and your network would break.

OK that concludes this blog entry.

Thanks Pedro for the question.

Please feel free to send me questions and I'll do my best to answer all of them.

I'll see you tomorrow.

All the very best.