OSPF router-id

nel

Quick one about the router-id. In the self study guide it says that using the router id command is the preferred method and overrides the loopback and physical interfaces. but if the router is rebooted or looses power then the router id is reset. So if a loopback is created as the router id then its more stable as its always up aslong as the device is functioning.

So why is one recommended even though the other is more stable? is it because if you need to clear it you only have to clear the ospf process?

Its clear the positives and negatives about both techniques but i just dont feel the author gets to the point.

Sorry for being a dumb ass but i like to get a full picture.

mikearama

If the router reboots the ID is lost? I don't believe that's so... perhaps only if you don't issue the wr mem or copy run start. When that's done, it's stored in the config.

tech-airman

nel wrote:

Quick one about the router-id. In the self study guide it says that using the router id command is the preferred method and overrides the loopback and physical interfaces. but if the router is rebooted or looses power then the router id is reset. So if a loopback is created as the router id then its more stable as its always up aslong as the device is functioning.

So why is one recommended even though the other is more stable? is it because if you need to clear it you only have to clear the ospf process?

Its clear the positives and negatives about both techniques but i just dont feel the author gets to the point.

Sorry for being a dumb ass but i like to get a full picture.

nel,

A general pattern that I'm noticing about studying for the BSCI exam is that there's a scale, from low to high, of fully automated to manual overriding of automation. For example, by default (or automatically) OSPF chooses the highest IP address of all of the physical interfaces as the router id. But what if that physical interface goes down? So you have to manually override OSPF's tendency to choose the highest IP address of all of the physical interfaces by configuring a loopback interface so that OSPF will tend to automatically use the IP address of the loopback address to be the router id. I guess a reason why you'd want to manually configure the OSPF router id is in the case you've got multiple loopback interfaces configured on that router and the highest IP address among the loopback interfaces is NOT intended for OSPF to choose as it's router ID. Unfortunately, I can't come up with an example of why you'd want to manually override OSPF's decision of highest physical interface IP address then highest loopback interface IP address, but as an implementor, I'd imagine it would be useful to know that it's possible to manually override OSPF's automatic router id choosing process if you ever need to.

nel

mikearama wrote:

If the router reboots the ID is lost? I don't believe that's so... perhaps only if you don't issue the wr mem or copy run start. When that's done, it's stored in the config.

Thats what i thought but i cant test it until i get home. but on page 186 it states "after the router id has been set it does not change even if the interface that the router is using goes down. the router id changes only if the router id reloads or if the ospf routing process restart"

nel

tech-airman wrote:

nel,

A general pattern that I'm noticing about studying for the BSCI exam is that there's a scale, from low to high, of fully automated to manual overriding of automation. For example, by default (or automatically) OSPF chooses the highest IP address of all of the physical interfaces as the router id. But what if that physical interface goes down? So you have to manually override OSPF's tendency to choose the highest IP address of all of the physical interfaces by configuring a loopback interface so that OSPF will tend to automatically use the IP address of the loopback address to be the router id. I guess a reason why you'd want to manually configure the OSPF router id is in the case you've got multiple loopback interfaces configured on that router and the highest IP address among the loopback interfaces is NOT intended for OSPF to choose as it's router ID. Unfortunately, I can't come up with an example of why you'd want to manually override OSPF's decision of highest physical interface IP address then highest loopback interface IP address, but as an implementor, I'd imagine it would be useful to know that it's possible to manually override OSPF's automatic router id choosing process if you ever need to.

Hi Mate,

I understand all the factors which can effect the choice of what becomes the router ID but i dont understand what the author is trying to say. Like i said she recommends the router id command but then goes on to state the loopback is more stable and that the rid is cleared on reboot or by clearing the process? even at my ccna studies i used to lab with the router id command and use one of the outgoing interface IP's and im sure its written in the config but the quote from the book seems to contradict what im thinking. unless its been put across in the book in the wrong context perhaps?

kryolla

The router id command is recommended more the using the loopback with the highest IP because when you peer with virtual links you use the RID and if someone comes behind you and put another loopback with a higher IP and the process is re-started you just lost your virtual link. So it is recommend to hard code the router id. HTH

EdTheLad

By default on startup ospf will choose the highest interface ip address as its router-id, if this interface fails ospf will still use this ip address.If the router reloads and the interface is still down after reload, ospf will still use this interface as the router-id.
If a higher ip address was present after reload this address will be used after reload.

As kryolla mentions if your RID changes your virtual links are screwed, if you are using virtual links in the first place your ospf network is screwed as this is a terrible idea.Cant remember off hand if sham links use the router-id.

For consistency create a loopback address and use this as the router-id, end of story.

kryolla

also look into OSPF RID and BGP RID being the same in order for redistribution but dont quote me on it.