[j-nsp] RE : RPD soft assertion failed

[Daniel Verlouw]

Hi David,

On May 27, 2009, at 9:18 PM, <david.roy at orange-ftgroup.com> <david.roy at orange-ftgroup.com 
 > wrote:
> Do you have some configuration at this level "edit protocols bgp  
> path-selection" ?

no, it's empty.

> Did the RPD restart ? It seems that yes : "%KERN-6: pid 12767  
> (rpd),uid 0: exited on signal 6 (core dumped)

no, rpd actually does not restart:

daniel at jun1.bit-1> show system processes detail wide | match rpd
  1149     0     1   0   4  0 410136 kqread  3:02AM  ??  S      
97:48.47 /usr/sbin/rpd -N

daniel at jun1.bit-1> show system uptime
Current time: 2009-05-27 21:56:30 CEST
System booted: 2009-05-27 03:01:52 CEST (18:54:38 ago)

in fact, there appears to be no impact at all besides a core dump  
being generated. We don't see any dropped adjacencies etc.

> Please, give us feedback regarding to the root cause, when you will  
> have more info? We've planned to use 9.3R3.8, too ;-)

I'll definitely follow-up on-list once we have some more news from  
JTAC. Oddly enough, up until now, every prefix being logged along with  
each coredump is originated from Eastern Europe/Russia, AS9198 being  
the "top talker".

    --Daniel.