cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1363
Views
5
Helpful
3
Replies

RIPv2 Holddown Timer

Craddockc
Level 3
Level 3

Community,

I am labbing up RIPv2 in GNS3 and came across something interesting. The CCIE R/S v5.0 book states the following about the Holddown timer in RIPv2:

"Router starts the Holddown timer for this network. While the timer runs (180 seconds by default), the router itself advertises the network with infinite metric (Route Poisoning) to force its neighbors to find an alternative route if possible. Additionally, the router locks the routing entry in its routing table, still pointing toward the former next hop. Absolutely no updates whatsoever are accepted until the Holddown timer expires, regardless of who sends them and what metric they claim."

 

This passage implies that the router will advertise the invalidated route with an infinite metric for the duration of the Holddown timer. However, I have found this to not be true in my lab. I have found that the router will continue to advertise the invalidated route as unreachable for (Flush timer - invalid timer). Essentially the router will continue to advertise the route as invalid for the remaining period on the Flush timer and even AFTER the Holddown timer has expired unless it receives an update about the invalidated route between the Holddown timers expiry and the Flush timers expiry. So it would seem the Holddown timer does not really affect how long the router advertises the route as inaccessible. However, if the router does receive a valid update about the inaccessible route after the Holddown expires but before the Flush expires, it will send a flash update regarding that and stop advertising the route as unreachable.

 

Can anyone corroborate this finding? Thanks!

 

1 Accepted Solution

Accepted Solutions

Hello Craddockc,

>> So it would seem the Holddown timer does not really affect how long the router advertises the route as inaccessible

 

When the route is in holddown state the route is reported in IP routing table as possibly down and no update should be accepted before holddown timer expires or the route is removed from the routing table.

The holddown timer refers to this aspect and not to the fact of advertising the route with infinite metric = 16 in RIP.

 

As reported  in both books there are cases there the holddown timer will not be allowed to expire (180 seconds) for the simple reason that another timer will remove the route before.

The idea is that in a simple distance vector routing protocol if a topology change happens, the router cannot accept other routers updates about the route for some time, because updates are sent every 30 seconds so it is necessary to wait that all routers are aware of the change noticed by the local router and signalled with poison reverse.

 

In any case, RIPv2 convergence is still too much slow for modern networks and it is not widely used.

 

Hope to help

Giuseppe

 

View solution in original post

3 Replies 3

Craddockc
Level 3
Level 3

Community,

I think I found the answer to this question. Chapter 5 of the Routing TCP/IP vol1  book by Jeff Doyle states the following:

"The route will be advertised with the unreachable metric until the garbage collection timer expires, at which time the route will be removed from the route table"

This statement is in alignment with what I was seeing in my lab. 

Hello Craddockc,

>> So it would seem the Holddown timer does not really affect how long the router advertises the route as inaccessible

 

When the route is in holddown state the route is reported in IP routing table as possibly down and no update should be accepted before holddown timer expires or the route is removed from the routing table.

The holddown timer refers to this aspect and not to the fact of advertising the route with infinite metric = 16 in RIP.

 

As reported  in both books there are cases there the holddown timer will not be allowed to expire (180 seconds) for the simple reason that another timer will remove the route before.

The idea is that in a simple distance vector routing protocol if a topology change happens, the router cannot accept other routers updates about the route for some time, because updates are sent every 30 seconds so it is necessary to wait that all routers are aware of the change noticed by the local router and signalled with poison reverse.

 

In any case, RIPv2 convergence is still too much slow for modern networks and it is not widely used.

 

Hope to help

Giuseppe

 

Thank you Giuseppe for your detailed response! This makes alot of sense!

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community:

Innovations in Cisco Full Stack Observability - A new webinar from Cisco