All things related to multicast

[Amel Caldwell <>]

Brent--

As you know I have been seeing this as well with you and within our exchange
point.  It looks like the problem started for me at May 3, 08:13:10 PDT.

May  3 08:13:10 hnsp1-wes rpd[548]: RPD_MSDP_PEER_DOWN: MSDP peer 198.48.91.78
peer-group none out of Established
May  3 08:13:46 hnsp1-wes rpd[548]: RPD_MSDP_PEER_DOWN: MSDP peer 198.48.91.78
peer-group none out of Established
May  3 08:36:28 hnsp1-wes rpd[548]: RPD_MSDP_PEER_DOWN: MSDP peer 198.48.91.78
peer-group none out of Established
May  3 09:17:29 hnsp1-wes rpd[548]: RPD_MSDP_PEER_DOWN: MSDP peer 198.48.91.78
peer-group none out of Established
May  3 09:22:14 hnsp1-wes rpd[548]: RPD_MSDP_PEER_DOWN: MSDP peer 198.48.91.78
peer-group none out of Established
May  3 10:42:07 hnsp1-wes rpd[548]: RPD_MSDP_PEER_DOWN: MSDP peer 198.48.91.78
peer-group none out of Established
May  3 10:51:04 hnsp1-wes rpd[548]: RPD_MSDP_PEER_DOWN: MSDP peer 198.48.91.78
peer-group none out of Established

There is nothing else logged nearby in any of my logs.

Anyone know of anything that happened at this time?

Amel

On Fri, 10 May 2002, Brent Sweeny wrote:

>folks, for the last several days we (Abilene) are seeing frequent--and
>even weirder--simultaneous resets of msdp sessions on Abilene, and
>can't quite figure out what's going wrong.  here are some data points,
>and I'll ask some questions at the end.
>-- they SEEM to be happening just between the Abilene GSRs and various
>   peers who all appear to be Junipers
>-- we're seeing this sort of message logged simultaneously across the
>   entire Abilene network, from the east coast and west (19 msdp peers
>   fail on it at the same second; probably another 30 or so stay up):
>  May 10 20:50:42 UTC: %MSDP-5-PEER_UPDOWN: Session to peer 192.12.206.250 
> going down
>  May 10 20:50:42 UTC: MSDP: 192.12.206.250: Peer reset, other side reset
>-- it's happening pretty much constantly.  the most they stay up between
>   resets is about 15 mins.
>-- our working theory is that Abilene is picking up something and resending
>   it to everyone else that makes some peers barf. but we can't figure out 
> what,
>   and if so which side is broken.
>-- this doesn't seem to be a GENERAL cisco-juniper interaction--others are
>   working just fine--so it must be something about our traffic.  of course 
> it
>   may not be a cisco-juniper thing at all...
>-- we've debugged cisco msdp detail and resets and messages and don't see
>   anything vaguely related.  it all looks normal as far as we can tell.
>-- we've also debugged one juniper side ("msdp packet") and see nothing at
>   all.  it doesn't even say the session went down. otherwise its traffic
>   looks normal (as far as we can tell): normal SAs being exchanged etc.
>-- we've not yet opened a case with cisco or juniper but probably will.
>
>esp since many of you also have Junipers which peer with Abilene, if you
>see anything useful, or can think of something else specific to look at,
>or better yet KNOW of something, we'd be grateful.
>   TIA,        brent
>--
>Brent Sweeny            
>
>            812 / 855-5391
>Indiana University Information Technology/Abilene Network Operations
>2711 E 10th Street, Bloomington, IN  47408
>PGP fingerprint = 13 6C E6 F9 70 7A 89 D6  7E D3 9E DD 68 A2 69 44
>
>