Great thank Partha. I think replacing quartz-1.6.1.jar with quartz-1.8.0.jar fixed it.

Dear Sergej,

We are using Confluence 3.3 and facing the same problem. The automatic indexing fails, we have to flush queue manually.

That is likely not the same issue. You are most likely hitting the same issue as Steven IIg as described and which is also described in our knowledge base and the most likely cause of your out of memory error is daily xml backups.

Dear Royce,

1. If I already upgraded to 3.2.1_01, I just need to replace <tomcat>/webapps/confluence/WEB-INF/lib/quartz-1.6.1.jar with quartz-1.8.0.jar, then restart and do a full reindex via the admin. console.?

If you are using an exploded war (which I believe you are using tomcat) it is fine to just replace the jar and restart.

I also see the file: lucene.zip
Do I need to do anything with that zip file to fix this problem?

No that was a previous patch that is no longer needed. I will remove that from the ticket.

Kind Regards,
Partha Kamal

I am upgrading Confluence 2.10 to 3.2.1_01 via WAR file. Does it mean:
1. If I already upgraded to 3.2.1_01, I just need to replace <tomcat>/webapps/confluence/WEB-INF/lib/quartz-1.6.1.jar with quartz-1.8.0.jar, then restart and do a full reindex via the admin. console.?
2. I need to build the 3.2.1_01 WAR with quartz-1.8.0.jar and do my upgrade from 2.10 to 3.2.1_01 again?

I also see the file: lucene.zip 23/Apr/10 3:28 AM 7 kB Partha Kamal [Atlassian]
Do I need to do anything with that zip file to fix this problem?

Thanks.

We are using Confluence 3.3 and facing the same problem. The automatic indexing fails, we have to flush queue manually.

Please reopen this issue!

I have patched my Confluence 3.2 deployment and it has been working fine. We have a fairly large instance when attachments are included. I had left the daily backup process on even though we are backing up the database directly.
We had an instance where the daily backup failed with an OOM. Since that time other scheduled jobs are not running and I have to run them manually. Restarting the environment and disabling the daily xml backup solves this part of the problem.

We have 3.2.1_01 and this quartz update did not fix our problem.

Michael,

This bug affects Quartz 1.6.1, which shipped with Confluence 3.2. Earlier versions of Confluence should not be experiencing the same root cause for the indexing issues, and thus the patch might not address the problem. I'd recommend opening a support case.

Cheers,
Adam Laskowski

Hey,

we are using Confluence 3.0.0_01 and facing the same problem.
Is the attached file also working with our version?

Best regards,
Michael

Thanks for the reply. Our indexing froze. Flush did not work. Re-index did not work. It seemed to be stuck processing and required a restart to fix. Seems to be ok now, so, I will wait and see. Perhaps it is a different issue than discussed here.

Bob,

The Quartz library included in Confluence 3.1 is version 1.5.2. I'm not aware of any issues in that version similar to 1.6.1, which is the affected version for this bug, but you should be able to attempt the patch to see if it resolves your issue. If not, please raise a new support case and we'll assist there.

Cheers,
Adam Laskowski