Page 1 of 1

Indexer running at 100% for some time now

Posted: Tue Jan 22, 2008 1:34 pm
by mglenney
Scalix 11.1

Did a 'top' recently and noticed the indexer is running at 100% CPU. I've been watching it for a few days and it hasn't changed. Didn't really notice it before because it's an 8 core system and it's only tying up 1 core. I tried to shutdown Indexer but get this:

Code: Select all

[root@xdsmail ~]# omoff -d0 indexer
omoff : [OM 4904] Sub-system cannot be started, stopped or reset: 'indexer'
No subsystems shut down.


I did a search through the forums to try and figure out how to solve this but I'm not finding anything. Any ideas?

MG

Posted: Thu Jan 24, 2008 1:31 pm
by mglenney
Ok. No takers.

Well, does anyone know if it's safe to kill the process 'indexer'? This process has chewed up 529 hours of processor time and doesn't look like it's going to quit any time soon.

MG

Posted: Fri Jan 25, 2008 3:26 am
by Valerion
The indexer can't be stopped, as it's marked NON-STOP, unless you bring Scalix down.

I suggest you upgrade to 11.2.0 at least, the indexer there is running as a low-priority process. But it would also be good to see why it doesn't stop. Do you see anything in the logs?

Posted: Tue Mar 04, 2008 6:49 pm
by shambler
I can verify that this problem can occur in 11.2. It happened to me this morning. The Scalix Wiki has some instructions on how to identify what is going wrong. The resulting log file on my server contained same message repeating over and over again:

Code: Select all

[2008-03-05 09:20:24] Awake liveRequests=100
[2008-03-05 09:20:24] /var/opt/scalix/mn/s/indexwork/1204514403.0 still has 78 active requests and end found


I ended up restarting Scalix as the load had crept up to over 15 and the log file generating for the indexer had reached 800MB in a few minutes. I'm lucky I caught it when I did. I'm hoping that this issue is fixed in 11.3 as I'd hate for it to happen over the weekend or when I'm not around. Yikes!