-
Bug
-
Resolution: Fixed
-
High
-
5.3, 5.4
-
None
Problem:
Confluence stops working;
These are the messages found on the logs:
2013-12-09 09:20:30,155 ERROR [http-8090-35] [mywork.host.rest.ExceptionLogger] toResponse Exception thrown from resource -- url: /confluence/rest/mywork/latest/status/notification/count | userName: bjaison | referer: http://agile.ironmountain.com/confluence/pages/viewpage.action?spaceKey=rc&title=Calendar com.atlassian.activeobjects.internal.ActiveObjectsInitException: Active Objects failed to initalize for bundle com.atlassian.mywork.confluence-host-plugin
Caused by: java.sql.SQLIntegrityConstraintViolationException: ORA-00001: unique constraint (CONFLUENCE53A.U_AO_9412A1_AOUSER_USERNAME) violated
Workaround 1:
Disable the bundled plugin: Workbox - Host Plugin.
Workaround 2:
1. Shutdown Confluence
2. Backup your Confluence database
3. Run the following SQL queries on your Confluence database:
\\ This will remove the oauth dances your users have done with other external applications eg JIRA, and will mean your users will need to re-authenticate with those systems over oauth delete from AO_9412A1_USER_APP_LINK; \\ This will remove task order and whether notifications have been read or not delete from AO_9412A1_AOUSER;
4. Clear the plugin cache
5. Restart Confluence
Deleting the content of these two tables only removes the data mentioned above, and does not remove any actual Workbox entries, so this is a relatively low impact workaround for those who wish to keep the Workbox functionality.
![[Extranet] Page](/images/icons/generic_link_16.png "[Extranet] Page"){kind=icon}
[CONFSERVER-32046] Active Objects failed to initalize for bundle com.atlassian.mywork.confluence-host-plugin
Adhip Pokharel
added a comment - Hi Steve,
Thanks for the information. I appreciate your effort in trying to resolve this issue. Since June is almost around the corner, i think it's best for us to go ahead and plan our upgrade to 5.5.2 and not wait for this bug to be fixed. I guess we'll have to try workaround #2 again just like what we did last time.
Adhip Pokharel
added a comment - Hi Steve,
Thanks for the information. I appreciate your effort in trying to resolve this issue. Since June is almost around the corner, i think it's best for us to go ahead and plan our upgrade to 5.5.2 and not wait for this bug to be fixed. I guess we'll have to try workaround #2 again just like what we did last time. Hi Adhip
Thanks for getting in touch. We're currently working on this issue so I would like to think that we should have a solution to this issue soon. We're currently having some problems replicating the bug locally, but as soon as we have everything verified we'll start working on a fix. Please keep watching this ticket for updates.
Regards
Steve Haffenden
Confluence bugmaster
Atlassian
Adhip Pokharel
added a comment - Hi Steve,
Is there any possibility that this bug could be fixed on version 5.5.3 or later? I am trying to decide on a version that we can upgrade. Out of the 4 critical bugs for us, two of them have been resolved on 5.5.1. We can wait until the 1st week of June to upgrade our Sandbox before upgrading our Production instance on June 26. We need 2 weeks of time period upon completion of our Sandbox to test everything. If this bug is unlikely to be fixed by then, we would start to plan to upgrade to version 5.5.2. We are on version 5.4.3 at the moment. Thanks.
Adhip Pokharel
added a comment - Hi Steve,
Is there any possibility that this bug could be fixed on version 5.5.3 or later? I am trying to decide on a version that we can upgrade. Out of the 4 critical bugs for us, two of them have been resolved on 5.5.1. We can wait until the 1st week of June to upgrade our Sandbox before upgrading our Production instance on June 26. We need 2 weeks of time period upon completion of our Sandbox to test everything. If this bug is unlikely to be fixed by then, we would start to plan to upgrade to version 5.5.2. We are on version 5.4.3 at the moment. Thanks. Adhip Pokharel
added a comment - Thanks, Steve. It's good to know that this issue is currently in your backlog.
Adhip Pokharel
added a comment - Thanks, Steve. It's good to know that this issue is currently in your backlog. Hi Everyone
I just wanted to give you all a quick update on this issue as I know it's been affecting a number of you. As of May 2014 this issue is currently on our backlog and we will be investigating a fix in the coming weeks.
As we have more information on this issue we'll update this ticket. Thanks for being so patient with this.
Regards
Steve Haffenden
Confluence Bugmaster
Atlassian
Adhip Pokharel
added a comment - Hi Adrian,
We almost ran into the same issue but after about 10 minutes, we were able to start Confluence. We were almost ready to roll back as we had taken a DB back up right before we ran the query as mentioned in workaround #2. However, i lost some of of my personal tasks. Nothing major, so we decided to stay with the change. I don't want to do this every time we upgrade Confluence.
Adhip Pokharel
added a comment - Hi Adrian,
We almost ran into the same issue but after about 10 minutes, we were able to start Confluence. We were almost ready to roll back as we had taken a DB back up right before we ran the query as mentioned in workaround #2. However, i lost some of of my personal tasks. Nothing major, so we decided to stay with the change. I don't want to do this every time we upgrade Confluence. 
Adrián
added a comment - I can confirm In our case applying workaround #2 won't let Confluence start.
Adrián
added a comment - I can confirm In our case applying workaround #2 won't let Confluence start. Adhip Pokharel
added a comment - When is this bug going to be fixed? Are we still supposed to try one of the workaround every time we upgrade Confluence? We did workaround #2 when we upgraded to Confluence 5.4.3.
Looks like we need to do it again when we upgrade to 5.5 as this bug is not fixed yet.
Adhip Pokharel
added a comment - When is this bug going to be fixed? Are we still supposed to try one of the workaround every time we upgrade Confluence? We did workaround #2 when we upgraded to Confluence 5.4.3.
Looks like we need to do it again when we upgrade to 5.5 as this bug is not fixed yet. 
Ismael (Inactive)
added a comment - Hi Adrián,
The DB locks happened when the load on the DB got quite intense. On our TEST environment for example, where the load is lighter, I couldn't reproduce the DB lock issue.
Having said that, we have managed to fix the issue by running the queries in the ticket description, but I also had to delete from the "Notification" table:
delete from AO_9412A1_USER_APP_LINK; //this is from this ticket's description delete from AO_9412A1_AOUSER; //this is from this ticket's description delete from AO_9412A1_AONOTIFICATION; //had to also run this query. (!) This will remove all existing Notifications (!)
Then, I also had to clear the plugins cache
I have re-enabled the Workbox add-on and now, for the last 2 days, things have been back to normal.
Cheers,
Ismael (Inactive)
added a comment - Hi Adrián,
The DB locks happened when the load on the DB got quite intense. On our TEST environment for example, where the load is lighter, I couldn't reproduce the DB lock issue.
Having said that, we have managed to fix the issue by running the queries in the ticket description, but I also had to delete from the "Notification" table:
delete from AO_9412A1_USER_APP_LINK; // this is from this ticket's description
delete from AO_9412A1_AOUSER; // this is from this ticket's description
delete from AO_9412A1_AONOTIFICATION; //had to also run this query. (!) This will remove all existing Notifications (!)
Then, I also had to clear the plugins cache
I have re-enabled the Workbox add-on and now, for the last 2 days, things have been back to normal.
Cheers,
We still get this error in the log in 5.5.3, almost everywhere, even when listing user directories.