FreeNAS components crashing

[Bidule0hm]

Sig 11 is a segmentation fault so yep, a program who ask access to the OS for some part of RAM which is owned by another program (for security reasons a program can't access to the memory reserved by another program). It's usually a development error but in this case I bet it was caused by some corruption of the boot drive and the reinstall took care of it.

 

[GoranTheGreat]

I'm seeing a variant of this issue. Exactly one (1) process exits with signal 11 each day:

Code:

[root@storage /var/log]# grep "exited on signal 11" /var/log/messages
Aug  8 17:16:54 storage kernel: pid 76704 (php), uid 80: exited on signal 11
Aug  9 17:31:53 storage kernel: pid 24976 (php), uid 80: exited on signal 11
Aug 10 17:32:07 storage kernel: pid 73552 (php), uid 80: exited on signal 11
Aug 11 17:46:53 storage kernel: pid 20071 (php), uid 80: exited on signal 11
Aug 12 18:02:10 storage kernel: pid 66512 (php), uid 80: exited on signal 11
Aug 13 18:16:53 storage kernel: pid 13105 (php), uid 80: exited on signal 11
Aug 14 18:31:53 storage kernel: pid 60165 (php), uid 80: exited on signal 11
Aug 15 18:33:12 storage kernel: pid 8336 (php), uid 80: exited on signal 11
Aug 16 18:33:13 storage kernel: pid 55335 (php), uid 80: exited on signal 11

Interesting that it's been within a second the last two days; and it always happens later than it did the day before.

There appears to be nothing else in messages near these dates/times:

Code:

[root@storage /var/log]# tail messages
Aug 15 12:35:23 storage uwsgi: [freeadmin.navtree:562] Couldn't retrieve http://192.168.1.5/plugins/owncloud/1/_s/treemenu: timed out
Aug 15 13:13:19 storage uwsgi: [freeadmin.navtree:562] Couldn't retrieve http://192.168.1.5/plugins/owncloud/1/_s/treemenu: timed out
Aug 15 13:15:13 storage uwsgi: [freeadmin.navtree:562] Couldn't retrieve http://192.168.1.5/plugins/owncloud/1/_s/treemenu: timed out
Aug 15 15:24:40 storage uwsgi: [freeadmin.navtree:562] Couldn't retrieve http://192.168.1.5/plugins/owncloud/1/_s/treemenu: timed out
Aug 15 16:55:08 storage uwsgi: [freeadmin.navtree:562] Couldn't retrieve http://192.168.1.5/plugins/owncloud/1/_s/treemenu: timed out
Aug 15 18:33:12 storage kernel: pid 8336 (php), uid 80: exited on signal 11
Aug 16 00:00:00 storage syslog-ng[20609]: Configuration reload request received, reloading configuration;
Aug 16 09:58:01 storage uwsgi: [freeadmin.navtree:562] Couldn't retrieve http://192.168.1.5/plugins/owncloud/1/_s/treemenu: timed out
Aug 16 18:33:13 storage kernel: pid 55335 (php), uid 80: exited on signal 11
Aug 17 00:00:19 storage syslog-ng[20609]: Configuration reload request received, reloading configuration;

 

[GoranTheGreat]

Yesterday today it took until 18:47 to crap out with signal 11...

Code:

...
Aug 15 18:33:12 storage kernel: pid 8336 (php), uid 80: exited on signal 11
Aug 16 18:33:13 storage kernel: pid 55335 (php), uid 80: exited on signal 11
Aug 17 18:47:14 storage kernel: pid 3732 (php), uid 80: exited on signal 11

What can this be about? Some process owned by uid 80, segfaulting after 24+ hours. The one on Aug 8 is actually the first one, there are no signal 11's in the older logs. This installation isn't very old, I set it up somewhere back in July.

 

[danb35]

I'm 99% sure that php isn't used in a standard FreeNAS installation, which would make it a jail that was using it. uid 80 is www, which makes sense.

 

[GoranTheGreat]

Thanks. The only jail I have so far is for the Owncloud plugin. Will try to disable that for a while.

Edit: Indeed, it is from the owncloud jail:

Code:

[root@storage /mnt/tank/jails/owncloud_1/var/log]# tail dmesg.today 
pid 24976 (php), uid 80: exited on signal 11
nfsd: can't register svc name
pid 73552 (php), uid 80: exited on signal 11
pid 20071 (php), uid 80: exited on signal 11
pid 66512 (php), uid 80: exited on signal 11
pid 13105 (php), uid 80: exited on signal 11
pid 60165 (php), uid 80: exited on signal 11
pid 8336 (php), uid 80: exited on signal 11
pid 55335 (php), uid 80: exited on signal 11
pid 3732 (php), uid 80: exited on signal 11

 

[GoranTheGreat]

I think it must be this cron job inside the owncloud jail that runs as www every 15 minutes:

Code:

root@owncloud_1:/ # crontab -l -u www
*/15 * * * * /usr/pbi/owncloud-amd64/bin/php -f /usr/pbi/owncloud-amd64/www/owncloud/cron.php

 

[DrKK]

PHP crashing is standard behavior. NTBF. :)