[j-nsp] Re: BGP neighbor state flapping

[Josef Buchsteiner]

Thursday, June 2, 2005, 9:46:09 AM, you wrote:

   
   
>>
 >> On Wed, Jun 01, 2005 at 04:35:31PM +0200, Hahn, Ch wrote:
 >> > After some minutes it's started again with "trace_on:
 >> Tracing to "/var/log/xxxx" started"
 >> >
 >> > Could it be that we had an daemon problem on the box? Is
 >> the above log entry indicating that the bgp process was restarted?
 >>
 >> Unfortunately there is no dedicated BGP process, so it looks
 >> like RPD restarted indeed. Check:
 >> show system processes wide detail | match "(STARTED|/rpd)"
 >> and you should see when the rpd fired up the last time.
 
HC>  Thanks, that gave me a hint. I checked the messages log and
HC> found that indeed the rpd was restarted several times and this
HC> correlates with the log entries in the bgp trace logs. It seems
HC> that the restart was due to a page read fault for the everytime
HC> same pc addresses. Am I right that this could indicates an
HC> corrupted memory modul?


    just   open   a   case   with   juniper  support  and  upload  the
    /var/tmp/rpd.core.*  file  to  ftp.juniper.net/pub/incoming  along
    with the message log and you will get an answer in a few minutes [
    one  we  have  the  core  file  ] and most likely the solution and
    possible workaround too...

    Josef
 
HC>  cheers,
HC>  Christian
 >>
 >> Do you run any other protocols on that box? Do you see
 >> similar adjacency flaps there? If you do run others but see
 >> no flaps, it's not the RPD restarting...
 >>
 >>
 >> Best regards,
 >> Daniel
 >>
 >> --
 >> CLUE-RIPE -- Jabber: dr at cluenet.de -- dr at IRCnet -- PGP:
 >> 0xA85C8AA0 _______________________________________________
 >> juniper-nsp mailing list juniper-nsp at puck.nether.net
 >> http://puck.nether.net/mailman/listinfo/juniper-nsp
 >>
HC>  _______________________________________________
HC>  juniper-nsp mailing list juniper-nsp at puck.nether.net
HC>  http://puck.nether.net/mailman/listinfo/juniper-nsp