-
A [small change to the Timer in RemoveUnattachedAssetsScheduler][1] seems to have fixed the problem with the ecdc environment.
[1]: https://fedorahosted.org/aplaws/changeset/2022.
2009-10-12 10:37:23 UTC by terry_permeance
-
For what it's worth, ecdc now has a "stop" target for Tomcat as of [r2021][1].
Usage: ant stop
But as discuss below, Tomcat shuts down but the process doesn't stop.
[1]: https://fedorahosted.org/aplaws/changeset/2021.
2009-10-12 09:28:49 UTC by terry_permeance
-
Thanks for the information. Didn't know these options. I think we use it for the 1.0.5 release and find a solution to stop the threads gracefully for the next release.
Thanks
Peter.
2009-10-08 06:49:02 UTC by pboy
-
Hi,
As the title reads, I've discussed with pboy about implemeting a freeze on the 1.0.5 tree, to initiate the process of releasing the next version of APLAWS.
Camden would like perform some testing on the software prior to release, therefore it would be of interest to freeze shortly. Currently, the tree builds cleanly, with most recent patches to ccm-ldn-* and multilanguage by Terry and...
2009-10-05 18:12:12 UTC by pfosterlbc
-
It's only a workaround, but you can set $CATALINA_PID environment variable to a file location. When tomcat starts it writes the process id to this file. Then you can use
"catalina stop -force"
which will forcibly kill the process.
Could also try getting a thread dump (I think "kill -QUIT" does this) to see if it indicates which thread is stuck...
2009-09-30 14:09:19 UTC by britsp
-
One of our guys did some work looking at daemon vs non-daemon threads. I'll see if I can find the changes.
Terry.
2009-09-29 03:18:27 UTC by terry_permeance
-
I'll ask one of our guys if do anything different, because it works for us.
2009-09-28 14:15:20 UTC by shawnlane
-
To look for deadlocked process in Postgres:
It looks as though the deadlock processes now return to the normal 'idle in transaction' state after 30 minutes. If this doesn't happen you can use the following commands to manually kill the process:-
Log on to the appropriate machine and run this command:
grep deadlock /var/log/postgresql.log
(you may need to be root,)
NB: If the file...
2009-09-28 14:13:36 UTC by shawnlane
-
ccm stop command does not work correctly, at least in Tomcat and Linux. It stops ccm applications, but doesn't finish the Tomcat process. I put some information on trac:
https://fedorahosted.org/aplaws/ticket/27
Please, could someone who is running APLAWS in a windows environment wether Windows is affected as well?
I think, there are 3 ways to handle it:
- add a shutdown...
2009-09-28 06:14:48 UTC by pboy
-
This issue seems not to be resolved yet.
Shawnlane: Please could you describe how you monitor deadlocks in your postgresql environment? Perhaps we could prepare a workaround?
Alan: Could you test your solutions? Are they ready to be commited to trunk?
Regards
Peter.
2009-09-28 06:03:19 UTC by pboy
