Routing Protocol Question I cant understand

The OSPF dead timer is 3 minutes. But if I switch off a router for POC testing, the switchover is almost instant. Can anyone explain why?

Find more posts tagged with

Comments

networker050184

When the interface goes physically down the neighbor is removed immediately.

dppagc

Oh I didnt know that. Thanks.
In that case what is the dead timer for?
Because they are not directly connected?
Is it possible to illustrate with an example?

OctalDump

I should know the answer to this, but I'm only 80% sure it is for situations where the link is up but the router is not getting hellos for whatever reason - eg intermediate link is down (for example where there is an intermediate switch), interface has been configured passive, OSPF process has failed, overloaded router, etc.

So it could still have connectivity, pings etc, but OSPF is not working or has had configuration changed.

james43026

Yeah, the dead timer basically does one thing. It allows the router to say to itself, if I haven't heard a hello from you within the constraints of the dead timer, then I'm going to consider you to be down.

dppagc

But if a router goes down, wouldnt its neighboring routers signal to other routers that it is down?
Isnt it faster than waiting for the dead interval?

Sy Kosys

dppagc wrote: »

But if a router goes down, wouldnt its neighboring routers signal to other routers that it is down?
Isnt it faster than waiting for the dead interval?

That's what the hello timer is for. With no hello packets being sent by R1 to R2, R2 will wait the dead interval specified before considering the R1 neighbor down.

dppagc

but the assumption is that R1 and R2 are not directly connected right?