GO HERE Seg Fault ??

[atsak]

This is a PIAF 2.0.6.5 system on an IBM x3650 m4 quad core with RAID1 / 4GB RAM. Freepbx 2.11 and Asterisk 11.9.


Jul 1 09:04:45 servername kernel: asterisk[10786]: segfault at 8 ip 0026829e sp b054a9a0 error 4 in libc-2.12.so[1f4000+191000]
Twice this morning, within 5 minutes.

System under some load; two conferences going on one with about 20 users, the other with 3. Otherwise a slow day phone wise; holiday up here (happy Canada Day).

Can't find any clues on this particular issue. Wondered if it's a continuation of the unistim seg fault that existed in 11.7 but can't see any indicators of that, and unistim phones weren't in use at the time.

Thoughts?

 

[billsimon]

"I'm feeling lucky": http://pbxinaflash.com/community/index.php?threads/sqlite2-in-piaf.14864/#post-95573

 

[atsak]

Ha ya I thought of that but those commands were already in (and confirm the modules noload)

 

[atsak]

Still happening - seems to be caused by conferencing . . . how can I trace the cause of it - I don't know enough about Linux to figure that part out?

 

[wardmundy]

Might be worth upgrading Asterisk just to see if it goes away.

 

[atsak]

I'm going to at 10 PM to 11.10 (using your handy little script but modifying it for 11.10.2 which I think was recently released). Guess we'll see.

 

[Capstone Works]

Status on your results? We went to 11.10.2 and still have a very similar issue.

 

[atsak]

Nope, still happens - very rare though (once a month). Is definitely tied to conferencing, but can't replicate it with any regularity for sure; after the seg fault it just starts right up. Weirdly seems to happen with inbound calls from a lot of cell phones; could be coincidental.

 

[Capstone Works]

I think it's related to the sqllite bug. We made this change ( link in next post) and things immediately improved. I'll report back in a few days.

 

[Capstone Works]

http://pbxinaflash.com/community/index.php?posts/100122/