[Spacewalk-list] Taskomatic crashes during repodata generation

[Jeremy Maes]

guirimand, michael schreef op 03/11/2014 09:21:
>
> Hi Jeremy,
>
> We're expecting an issue closed to this one.
>
> It was related to JVM configuration.
>
> You need to change --Xms --Xmx.
>
> https://access.redhat.com/solutions/43122
>
Hey

I should've mentioned that I've changed the taskomatic memory a long 
time ago. That seemed to have stopped the crashed back then, though it 
was before I added the bigger repositories.

Current values are: (in /etc/rhn/rhn.conf as the newer spacewalk 
versions demand)
> taskomatic.java.initmemory=256
> taskomatic.java.maxmemory=2048

I could of course increase this further, but I'm mostly wondering why 
I'm not seeing any OutOfMemory errors anywhere. Not in taskomatic logs, 
not in tomcat logs. Everyone having that specific issue seems to mention 
those errors, just as the RedHat KB entry says. I did however notice 
that my /etc/sysconfig/tomcat6 file had had the original line with max 
256MB memory appended again, probably overwriting my earlier change to 
512MB. Will change it back to 512 and see how it goes.

Regards,
Jeremy

>
> *Michael Guirimand*
>
> System Administrator
>
> *From:*spacewalk-list-bounces at redhat.com 
> [mailto:spacewalk-list-bounces at redhat.com] *On Behalf Of *Jeremy Maes
> *Sent:* Monday 3 November 2014 09:10
> *To:* spacewalk-list at redhat.com
> *Subject:* [Spacewalk-list] Taskomatic crashes during repodata generation
>
> Hey
>
> Recently I've started syncing a couple of bigger repositories (RHEL, 
> CentOS, ...) and since then my taskomatic will almost systematically 
> crash every time it gets to the "Generating new repository metadata 
> for channel" stage. When it does, it usually seems to crash very fast 
> after starting this task for one of the bigger channels.
>
> I've already enabled debug logging, but the only errors I'm getting are:
> */var/log/rhn/rhn_taskomatic_daemon.log
>
> *
>
> ERROR  | wrapper  | 2014/10/31 21:46:46 | JVM appears hung: Timed out 
> waiting for signal from JVM.
> ERROR  | wrapper  | 2014/10/31 21:46:46 | JVM did not exit on request, 
> terminated
> STATUS | wrapper  | 2014/10/31 21:46:47 | JVM exited in response to 
> signal SIGKILL (9).
> ERROR  | wrapper  | 2014/10/31 21:46:47 | Unable to start a JVM
> STATUS | wrapper  | 2014/10/31 21:46:48 | <-- Wrapper Stopped
>
>
> */var/log/messages
>
> *
>
> Oct 31 21:46:46 spacewalk wrapper[22867]: JVM appears hung: Timed out 
> waiting for signal from JVM.
> Oct 31 21:46:46 spacewalk wrapper[22867]: JVM did not exit on request, 
> terminated
> Oct 31 21:46:47 spacewalk wrapper[22867]: JVM exited in response to 
> signal SIGKILL (9).
> Oct 31 21:46:47 spacewalk wrapper[22867]: Unable to start a JVM
> Oct 31 21:46:48 spacewalk wrapper[22867]: <-- Wrapper Stopped
>
> Apart from those I can't seem to find any other error messages (not in 
> the taskomatic log, not in syslog, not in tomcat logs, ...) at the 
> same time that could indicate what is going wrong.
>
> Anyone have any pointers as to where I could look next to hopefully 
> solve this problem? We're starting to use spacewalk more and more and 
> taskomatic always crashing is getting to be an issue. Of course I 
> could use a cronjob to check its status periodically and restart it if 
> necessary, but that's only a bandaid solution...
>
> -- 
> Jeremy Maes
>

 **** DISCLAIMER ****
http://www.schaubroeck.be/maildisclaimer.htm
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/spacewalk-list/attachments/20141103/e982bf78/attachment.htm>