[c-nsp] All RRs down

[Mark Tinka]

On Monday 10 May 2010 02:23:42 pm Nam NGUYEN wrote:

> I have a small network running IBGP with 2 designated
>  route reflectors. Everything now is working fine. My
>  question is that, if the 2 RRs are down for some
>  reasons, can other routers in the cluster keep
>  forwarding packets during the down time of RRs? In other
>  words, will routing table or forwarding table of a
>  router remain unchanged if its RRs go down?

If the route reflectors are down, the routers won't know of 
BGP routes after various timers expire. However, they should 
have some knowledge of IGP routes (I say "some" because your 
IGP configuration may or may not necessarily permit all 
routers having knowledge of all IGP routes).

You can use IETF Graceful Restart (if supported) between 
iBGP clients and their route reflectors so they can maintain 
forwarding state as the route reflectors restart (assuming 
they plan to restart), however, how long they can do so 
depends on the configured Graceful Restart timers. Also, any 
changes in the network during the restart period are not 
captured, but this may not be a huge problem as they might 
reflect a negligible % of your overall traffic being 
switched at the time if the route reflectors return to a 
functional state in a timely fashion.

Cheers,

Mark.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: This is a digitally signed message part.
URL: <https://puck.nether.net/pipermail/cisco-nsp/attachments/20100510/04fa63b3/attachment.bin>