[JBoss JIRA] (JBTM-1786) Change to TMFAIL for an ABORT on a non-prepared TX
by Tom Jenkinson (JIRA)
[ https://issues.jboss.org/browse/JBTM-1786?page=com.atlassian.jira.plugin.... ]
Tom Jenkinson updated JBTM-1786:
--------------------------------
Fix Version/s: 4.17.7
> Change to TMFAIL for an ABORT on a non-prepared TX
> --------------------------------------------------
>
> Key: JBTM-1786
> URL: https://issues.jboss.org/browse/JBTM-1786
> Project: JBoss Transaction Manager
> Issue Type: Bug
> Security Level: Public(Everyone can see)
> Components: JTA, JTS
> Affects Versions: 4.16.4
> Reporter: Johnathon Lee
> Assignee: Johnathon Lee
> Fix For: 4.17.7, 5.0.0.M4
>
>
> An ABORT on a non-prepared TX should be calling end with TMFAIL.
> It's more efficient in that the RM can return an RB* code immediately and short circuit the rollback internally if it likes. More importantly though the RM is less likely to object that it's still running tx activity on another thread and the concurrent TMSUCCESS therefore makes no sense to it.
> Also the XAResourceRecord.topLevelAbort code should immediately continue on to calling rollback on the resource despite the failed end().
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
12 years, 8 months
[JBoss JIRA] (JBTM-1786) Change to TMFAIL for an ABORT on a non-prepared TX
by Tom Jenkinson (JIRA)
[ https://issues.jboss.org/browse/JBTM-1786?page=com.atlassian.jira.plugin.... ]
Tom Jenkinson updated JBTM-1786:
--------------------------------
Fix Version/s: 5.0.0.M4
> Change to TMFAIL for an ABORT on a non-prepared TX
> --------------------------------------------------
>
> Key: JBTM-1786
> URL: https://issues.jboss.org/browse/JBTM-1786
> Project: JBoss Transaction Manager
> Issue Type: Bug
> Security Level: Public(Everyone can see)
> Components: JTA, JTS
> Affects Versions: 4.16.4
> Reporter: Johnathon Lee
> Assignee: Johnathon Lee
> Fix For: 5.0.0.M4
>
>
> An ABORT on a non-prepared TX should be calling end with TMFAIL.
> It's more efficient in that the RM can return an RB* code immediately and short circuit the rollback internally if it likes. More importantly though the RM is less likely to object that it's still running tx activity on another thread and the concurrent TMSUCCESS therefore makes no sense to it.
> Also the XAResourceRecord.topLevelAbort code should immediately continue on to calling rollback on the resource despite the failed end().
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
12 years, 8 months
[JBoss JIRA] (JBTM-1712) perf problem in FileSystemStore.openAndLock
by RH Bugzilla Integration (JIRA)
[ https://issues.jboss.org/browse/JBTM-1712?page=com.atlassian.jira.plugin.... ]
RH Bugzilla Integration commented on JBTM-1712:
-----------------------------------------------
tom.jenkinson(a)redhat.com changed the Status of [bug 968125|https://bugzilla.redhat.com/show_bug.cgi?id=968125] from ASSIGNED to POST
> perf problem in FileSystemStore.openAndLock
> -------------------------------------------
>
> Key: JBTM-1712
> URL: https://issues.jboss.org/browse/JBTM-1712
> Project: JBoss Transaction Manager
> Issue Type: Enhancement
> Security Level: Public(Everyone can see)
> Components: Transaction Core
> Affects Versions: 4.16.4, 5.0.0.M2
> Reporter: Jonathan Halliday
> Assignee: Mark Little
> Fix For: 4.17.6, 5.0.0.M3
>
>
> if(!file.exits())
> {
> if(createHierarchy(file))
> incorrectly calls the expensive (because synchronized) createHierarchy method in the common case where the file does not exist (because it's a uniq named new tx record) but the directory hierarchy does (because it's the create-once hashed dir tree). This causes excessive contention on the FileSystemStore instance object monitor lock. Should probably be something more like
> if(!file.getParent().exists())
> {
> if(createHierarchy(file))
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
12 years, 8 months