So, I can't remember what prompted me to test this but I came across an interesting issue.

we are pretty much taught that when you setup an OSPF instance you should set the router ID to an ip that exist on a loopback interface (in Mtik, that means using a bridge).
We are taught to do this because the interface never goes down (unless the router explodes or is turned off, mute point) and so the instance that is associating itself with the interface ip will not go down.

In my younger years saw first hand what happens when you don't use this.

So, using GNS3 and the new CHR router image I setup three routers with router IDs as 1.1.1.1 2.2.2.2 3.3.3.3
I setup bridges for those IPs as well. then constructed the interlink IPs and ospf network.
Next I turned off the bridge on one of the routers. I expected the OSPF to go down but it did not.
I tried a few things and in the end I have removed the bridges and their IPs and yet the OSPF still stays active.
I even rebooted to make sure.

So, this is the question, why is it staying up. what have I missed.



*Edit: I've just done the exact same thing on some Huaweis and I get the exact same result.
Gonna try some ciscos

*Edit: Cisco is the same

Hello

So, I can't remember what prompted me to test this but I came across an interesting issue.

we are pretty much taught that when you setup an OSPF instance you should set the router ID to an ip that exist on a loopback interface (in Mtik, that means using a bridge).
We are taught to do this because the interface never goes down (unless the router explodes or is turned off, mute point) and so the instance that is associating itself with the interface ip will not go down.

no there is nothing to do between ospf working and an ip on a loopback
most of the time you setup a loopback to be able to remote admin the router or setup ibgp, vpls etc....

In my younger years saw first hand what happens when you don't use this.

So, using GNS3 and the new CHR router image I setup three routers with router IDs as 1.1.1.1 2.2.2.2 3.3.3.3
I setup bridges for those IPs as well. then constructed the interlink IPs and ospf network.
Next I turned off the bridge on one of the routers. I expected the OSPF to go down but it did not.
I tried a few things and in the end I have removed the bridges and their IPs and yet the OSPF still stays active.
I even rebooted to make sure.

So, this is the question, why is it staying up. what have I missed.



*Edit: I've just done the exact same thing on some Huaweis and I get the exact same result.
Gonna try some ciscos

*Edit: Cisco is the same

this is standard behavior
as wrote above the loopback, nor the router-id is needed to make ospf working

aniway very strange you asked this because one question of my MTCNA or MTCRE ( i do not remember ) was

'what is the minimum setup needed to make ospf work" ?

this will answer your post

A+
Thierry

I have never come across this in any coarse material nor have I have had a question that poked a stick at this. I can't remember what it was but something prompted me to look into this, now I'm glad it did.

There is a lot of material out there that says setup a loopback for the ospf to auto select (again, ive seen what happens when it auto selects an int and it goes down). There is then other meterial that tells you how to set the router id, but never info that then tells that after you set the router id, that loopback that was talked about just a few sentences earlier is no longer needed.

Because of this I have met several people who were of my mind set.

Anyways, this was a nice bit of realisation