[infinispan-issues] [JBoss JIRA] (ISPN-2469) lock() producing timeout exceptions after nodes have left

[Mircea Markus (JIRA)]

     [ https://issues.jboss.org/browse/ISPN-2469?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mircea Markus updated ISPN-2469:
--------------------------------

    Fix Version/s: 5.2.0.Final

    
> lock() producing timeout exceptions after nodes have left
> ---------------------------------------------------------
>
>                 Key: ISPN-2469
>                 URL: https://issues.jboss.org/browse/ISPN-2469
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Locking and Concurrency, Transactions
>    Affects Versions: 5.2.0.Beta3
>            Reporter: Carsten Lohmann
>            Assignee: Mircea Markus
>            Priority: Critical
>             Fix For: 5.2.0.CR1, 5.2.0.Final
>
>         Attachments: LockAfterNodesLeftTest.java
>
>
> I have a replicated cache with locking mode set to PESSIMISTIC on which I use explicit locking before putting some value.
> After nodes have left the cluster, this does not work anymore and I get TimeoutExceptions like:
> {noformat}
> 2012-11-02 11:14:12,057 ERROR [InvocationContextInterceptor] (LockAfterNodesLeftTest.Putter-0) ISPN000136: Execution error
> org.infinispan.util.concurrent.TimeoutException: Could not acquire lock on key on behalf of transaction GlobalTransaction:<NodeA-31597>:4:local. Lock is being held by GlobalTransaction:<NodeA-31597>:4:local
> 	at org.infinispan.interceptors.locking.AbstractTxLockingInterceptor.newTimeoutException(AbstractTxLockingInterceptor.java:230)
> 	at org.infinispan.interceptors.locking.AbstractTxLockingInterceptor.waitForTransactionsToComplete(AbstractTxLockingInterceptor.java:210)
> 	at org.infinispan.interceptors.locking.AbstractTxLockingInterceptor.lockKeyAndCheckOwnership(AbstractTxLockingInterceptor.java:175)
> 	at org.infinispan.interceptors.locking.PessimisticLockingInterceptor.visitPutKeyValueCommand(PessimisticLockingInterceptor.java:120)
> {noformat}
> This can be reproduced with the attached unit test.
> (Use NUM_NODES_TO_STOP_FOR_TEST=0 to not stop any nodes - the test succeeds in that case)

--
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