[c-nsp] BGP Router CPU Utilization Sup720+iBGP+IPv6

[James P. Ashton]

Oddly enough,
 It seams to have run at 50% CPU for about an hour and is now back at 3% and seams to be stable there...

This concerns me a bit.  I don't see any errata regarding this for this IOS rev.

----- Original Message -----
From: "James P. Ashton" <james at gitflorida.com>
To: cisco-nsp at puck.nether.net
Sent: Thursday, October 8, 2009 11:52:38 AM GMT -05:00 US/Canada Eastern
Subject: [c-nsp] BGP Router CPU Utilization Sup720+iBGP+IPv6

Hello everyone,

I have just turned up iBGP between 2 7609s with Sup720-3BXLs 
(12.2(33)SRD1 AdIpServices)

I am now showing a constant 50% CPU use for the BGP Router process.

These routers have 450M of memory free.  I have 2 Full table IPv4 peers on each and an IPv4 iBGP session between them using Lo0.

I am attempting the same iBGP setup using IPv6.  


This is the iBGP config I have added to each router.  Very similar to what I have with IPv4.
======================

 neighbor internalv6 peer-group
 neighbor internalv6 remote-as 1***
 neighbor internalv6 description iBGP-peers
 neighbor internalv6 update-source Loopback0
 neighbor internalv6 version 4
!
 neighbor 2607:****:*:*::1 remote-as 1***
 neighbor 2607:****:*:*::1 peer-group internal
!
 address-family ipv6
  neighbor internalv6 send-community
  neighbor internalv6 next-hop-self
  neighbor internalv6 soft-reconfiguration inbound
  neighbor 2607:****:*:*::1 activate
 exit-address-family

=======================



Any thoughts on the CPU usage..?   I have one IPv6 session with a Level3 tunnel prior to this and had no CPU issues.   But am about to turn up another v6 link. This one on the second router.   Thats what prompted the iBGP turnup.

Thanks

James
_______________________________________________
cisco-nsp mailing list  cisco-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/