[infinispan-issues] [JBoss JIRA] (ISPN-4368) StateTransferOverwriteTest random failures

[Dan Berindei (JIRA)]

     [ https://issues.jboss.org/browse/ISPN-4368?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Dan Berindei updated ISPN-4368:
-------------------------------

              Status: Pull Request Sent  (was: Open)
    Git Pull Request: https://github.com/infinispan/infinispan/pull/2614


Ignore the CounterHashCode thread local in EquivalentConcurrentHashMapV8, it should not be considered a leak since it is static.

> StateTransferOverwriteTest random failures
> ------------------------------------------
>
>                 Key: ISPN-4368
>                 URL: https://issues.jboss.org/browse/ISPN-4368
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Core, Test Suite - Core
>    Affects Versions: 7.0.0.Alpha4
>            Reporter: Dan Berindei
>            Assignee: William Burns
>              Labels: testsuite_stability
>             Fix For: 7.0.0.Beta1
>
>
> {{testStateTransferInBetweenPrepareCommitWithPutCreate}} sometimes fails because the state transfer doesn't finish in time:
> {noformat}
> java.lang.RuntimeException: Timed out waiting for rebalancing to complete on node NodeBN-23267, current topology is CacheTopology{id=6, currentCH=DefaultConsistentHash{numSegments=60, numOwners=2, members=[NodeBM-52725, NodeBN-23267]}, pendingCH=DefaultConsistentHash{numSegments=60, numOwners=2, members=[NodeBM-52725, NodeBN-23267]}, unionCH=DefaultConsistentHash{numSegments=60, numOwners=2, members=[NodeBM-52725, NodeBN-23267]}}
> 	at org.infinispan.test.TestingUtil.waitForRehashToComplete(TestingUtil.java:214)
> 	at org.infinispan.distribution.rehash.BaseTxStateTransferOverwriteTest.doStateTransferInBetweenPrepareCommit(BaseTxStateTransferOverwriteTest.java:265)
> 	at org.infinispan.distribution.rehash.BaseTxStateTransferOverwriteTest.testStateTransferInBetweenPrepareCommitWithPutCreate(BaseTxStateTransferOverwriteTest.java:94)
> {noformat}
> The problem seems to be that the test intends to block the {{PUT_CREATE}} command on the primary owner of the test key, but it sometimes also blocks the state transfer put for the {{placeholder}} key - depending on which node was the backup owner of the {{placeholder}} key.
> The purpose of the {{placeholder}} key is also not very clear, we should add a comment how we expect the presence of another entry to interact with the tx / operation in progress.



--
This message was sent by Atlassian JIRA
(v6.2.3#6260)