[
https://issues.apache.org/jira/browse/DERBY-5082?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kristian Waagan updated DERBY-5082:
-----------------------------------
Attachment: derby-5082-1a-fix.diff
Attached patch 1a, which fixes the bug with a ShutdownException raised in
ContextManager by avoiding to destroy the tx if the database is shut down.
Also, if the transaction is being destroyed, catch ShutdownException in case
someone shuts down the database under our feet. This is already done in the
main processing loop if the worker thread detects that the daemon has been
stopped.
Since the patch is so small and isolated, I committed it right away with
revision 1076462.
The coming patch for DERBY-5088 will be larger, so I'll need to run more tests
before committing.
> ShutdownException in ContextManager.checkInterrupt() during shutdown
> --------------------------------------------------------------------
>
> Key: DERBY-5082
> URL: https://issues.apache.org/jira/browse/DERBY-5082
> Project: Derby
> Issue Type: Bug
> Components: Services
> Affects Versions: 10.8.0.0
> Reporter: Knut Anders Hatlen
> Attachments: derby-5082-1a-fix.diff, test.sql
>
>
> When running the attached repro script, I often (but not always) see a
> ShutdownException in ContextManager.checkInterrupt() when ij exits and shuts
> down the Derby engine. ij expects an exception here, but it expects an
> SQLException, not a ShutdownException.
--
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
