You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@geronimo.apache.org by "David Jencks (JIRA)" <ji...@apache.org> on 2010/10/17 21:47:22 UTC
[jira] Updated: (GERONIMO-5648) txmanager rollback needs retry
scheduler too
[ https://issues.apache.org/jira/browse/GERONIMO-5648?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
David Jencks updated GERONIMO-5648:
-----------------------------------
Attachment: GERONIMO-5648-1.patch
Initial patch idea. I haven't figure out how to write a test for it yet, and would like to review the logic changes.
> txmanager rollback needs retry scheduler too
> --------------------------------------------
>
> Key: GERONIMO-5648
> URL: https://issues.apache.org/jira/browse/GERONIMO-5648
> Project: Geronimo
> Issue Type: Bug
> Security Level: public(Regular issues)
> Components: transaction manager
> Affects Versions: 2.2, 3.0
> Reporter: David Jencks
> Assignee: David Jencks
> Fix For: 2.2.1, 3.0
>
> Attachments: GERONIMO-5648-1.patch
>
>
> The following sequence of events can result in a tx branch stuck in a RM until geronimo is restarted and recovers transactions:
> 1. tx branch prepares
> 2. geronimo decides to rollback the tx
> 3. RM crashes
> 4. attempt to rollback branch fails with e.g. a RETRY error code.
> We can do the same as we do for commits with rollbacks and keep trying until we succeed. I think we can also unify the two rollbackResources methods.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.