[infinispan-issues] [JBoss JIRA] (ISPN-3744) TopologyAwareTwoNodesMapReduceTest && SimpleTwoNodesMapReduceTest random failures

RH Bugzilla Integration (JIRA) issues at jboss.org
Mon Mar 17 08:26:10 EDT 2014


    [ https://issues.jboss.org/browse/ISPN-3744?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12953472#comment-12953472 ] 

RH Bugzilla Integration commented on ISPN-3744:
-----------------------------------------------

Vojtech Juranek <vjuranek at redhat.com> changed the Status of [bug 1075789|https://bugzilla.redhat.com/show_bug.cgi?id=1075789] from NEW to CLOSED
                
> TopologyAwareTwoNodesMapReduceTest && SimpleTwoNodesMapReduceTest random failures
> ---------------------------------------------------------------------------------
>
>                 Key: ISPN-3744
>                 URL: https://issues.jboss.org/browse/ISPN-3744
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Distributed Execution and Map/Reduce, Test Suite - Core
>    Affects Versions: 6.0.0.Final
>            Reporter: Dan Berindei
>            Assignee: Vladimir Blagojevic
>              Labels: testsuite_stability
>
> The static counter used in FailAfterSecondCallReducer in TopologyAwareTwoNodesMapReduceTest.testInvokeMapWithReduceExceptionPhaseInRemoteExecution and SimpleTwoNodesMapReduceTest.testInvokeMapWithReduceExceptionPhaseInRemoteExecution has several problems:
> 1. Both tests use the same counter, if they run in parallel one could reset the counter for the other.
> 2. The counter update is not atomic, so the async thread that executes the reducer on the originator and the remote executor thread that executes the reducer on the remote node can both see value 0.
> 3. There is no guarantee that the reducer is executed on the originator first, so the exception isn't always thrown on the remote node.

--
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


More information about the infinispan-issues mailing list