[JBoss JIRA] (JBTM-1511) Dynamic 1PC optimisation
by Mark Little (JIRA)
[ https://issues.jboss.org/browse/JBTM-1511?page=com.atlassian.jira.plugin.... ]
Mark Little resolved JBTM-1511.
-------------------------------
Resolution: Done
> Dynamic 1PC optimisation
> ------------------------
>
> Key: JBTM-1511
> URL: https://issues.jboss.org/browse/JBTM-1511
> Project: JBoss Transaction Manager
> Issue Type: Feature Request
> Security Level: Public(Everyone can see)
> Components: Transaction Core
> Affects Versions: 4.17.3
> Reporter: Mark Little
> Assignee: Mark Little
> Fix For: 5.0.0.M2
>
>
> If the first N-1 participants in the intentions list return read-only, then we can call commitOnePhase on the last resource, bypassing prepare and writing of the log.
> Note: this has been on the todo list for years. Time to do it.
--
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
11 years, 9 months
[JBoss JIRA] (JBTM-1336) DUPID detection for mysql in JTS mode does not work
by Mark Little (JIRA)
[ https://issues.jboss.org/browse/JBTM-1336?page=com.atlassian.jira.plugin.... ]
Mark Little commented on JBTM-1336:
-----------------------------------
How do you know that 0 will work for anything other than MySQL? We should get a fix driven into MySQL for a start and also see if we can bracket this test in our code with something that ensures it is only triggered when using a MySQL resource manager.
> DUPID detection for mysql in JTS mode does not work
> ---------------------------------------------------
>
> Key: JBTM-1336
> URL: https://issues.jboss.org/browse/JBTM-1336
> Project: JBoss Transaction Manager
> Issue Type: Bug
> Security Level: Public(Everyone can see)
> Components: JTS
> Reporter: Tom Jenkinson
> Assignee: Tom Jenkinson
> Priority: Minor
> Fix For: 5.0.0.M3
>
> Attachments: JBTM-1336.patch
>
> Original Estimate: 3 days
> Time Spent: 1 day
> Remaining Estimate: 2 days
>
> Mysql does not send errorcode of XAException.XAER_DUPID, it sends 0.
--
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
11 years, 9 months
[JBoss JIRA] (JBTM-1459) Create a parallel version of beforeCompletion invocations
by Mark Little (JIRA)
[ https://issues.jboss.org/browse/JBTM-1459?page=com.atlassian.jira.plugin.... ]
Mark Little commented on JBTM-1459:
-----------------------------------
If done then target TwoPhaseCoordinator and not the JTA transaction directly.
> Create a parallel version of beforeCompletion invocations
> ---------------------------------------------------------
>
> Key: JBTM-1459
> URL: https://issues.jboss.org/browse/JBTM-1459
> Project: JBoss Transaction Manager
> Issue Type: Enhancement
> Security Level: Public(Everyone can see)
> Components: JTA
> Affects Versions: 4.17.3
> Reporter: Michael Musgrove
> Assignee: Michael Musgrove
> Priority: Minor
> Fix For: 5.0.0.Final
>
> Original Estimate: 6 hours
> Remaining Estimate: 6 hours
>
> Some RM's do the bulk of there work in the beforeCompletion synchronisations so it would be beneficial to have an option for calling it asynchronously.
--
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
11 years, 9 months
[JBoss JIRA] (JBTM-1284) Implement concurrent nested transactions
by Mark Little (JIRA)
[ https://issues.jboss.org/browse/JBTM-1284?page=com.atlassian.jira.plugin.... ]
Mark Little commented on JBTM-1284:
-----------------------------------
Concurrent nested transactions are supported, so I'm not sure what the problem is.
> Implement concurrent nested transactions
> ----------------------------------------
>
> Key: JBTM-1284
> URL: https://issues.jboss.org/browse/JBTM-1284
> Project: JBoss Transaction Manager
> Issue Type: Feature Request
> Security Level: Public(Everyone can see)
> Components: Transaction Core
> Reporter: Tom Jenkinson
>
> Needs sanity checking
> We can collaborate on the work via linked GitHub branch
> The problem out of the box is when you AtomicAction "begin" - you will link to the ThreadActionData, as you can see in the test I can circumvent that via a different API call start passing in the parent directly.
> You can see there is precedent to manipulate ThreadActionData: ArjunaCore/arjuna/classes/com/arjuna/ats/arjuna/coordinator/AsyncPrepare.java (doing the prepare in multiple threads)
> I think this is feasible, it may need some thread safety via synchronization in the AtomicAction.
> Nested Transactions in core look to be able to support concurrent access except:
> 1. Threading may not work, will need to check and fix
> 2. ThreadActionData imposes a stack
> You can see in my test (linked in the GitHub branch) I am circumventing the stack.
> we already support nested transactions, it should just be not associating them with a thread and making sure they are thread safe when they talk to their parents
--
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
11 years, 9 months
[JBoss JIRA] (JBTM-1511) Dynamic 1PC optimisation
by Mark Little (JIRA)
Mark Little created JBTM-1511:
---------------------------------
Summary: Dynamic 1PC optimisation
Key: JBTM-1511
URL: https://issues.jboss.org/browse/JBTM-1511
Project: JBoss Transaction Manager
Issue Type: Feature Request
Security Level: Public (Everyone can see)
Components: Transaction Core
Affects Versions: 4.17.3
Reporter: Mark Little
Assignee: Mark Little
Fix For: 5.0.0.M2
If the first N-1 participants in the intentions list return read-only, then we can call commitOnePhase on the last resource, bypassing prepare and writing of the log.
Note: this has been on the todo list for years. Time to do it.
--
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
11 years, 9 months