[
https://issues.jboss.org/browse/ISPN-2622?page=com.atlassian.jira.plugin....
]
Dan Berindei commented on ISPN-2622:
------------------------------------
Looks like initial problem wasn't fixed: the coordinator is incrementing the RPC
counter for the final topology update only after all the members have installed that
topology, so it can happen after {{waitForRehashCompletion()}} finished.
Random failure in RpcManagerMBeanTest
-------------------------------------
Key: ISPN-2622
URL:
https://issues.jboss.org/browse/ISPN-2622
Project: Infinispan
Issue Type: Bug
Components: Core
Affects Versions: 5.2.0.Beta5
Reporter: Dan Berindei
Assignee: Dan Berindei
Priority: Minor
Labels: testsuite_stability
Fix For: 5.2.0.CR3
Attachments: testEnableJmxStats-0.log
The test records the replication counter in RpcManagerImpl at the start of the test and
after a put operation, and asserts that it only changed by 1. Occasionally, the assertion
fails:
{noformat}
java.lang.AssertionError: expected [1] but found [2]
at org.testng.Assert.fail(Assert.java:89)
at org.testng.Assert.failNotEquals(Assert.java:489)
at org.testng.Assert.assertEquals(Assert.java:118)
at org.testng.Assert.assertEquals(Assert.java:160)
at
org.infinispan.jmx.RpcManagerMBeanTest.testEnableJmxStats(RpcManagerMBeanTest.java:126)
{noformat}
This happens because replication counter is only incremented *after* a successful RPC.
The test starts when the cluster is considered "formed", that is when all the
nodes are included in the read CH on every node. But that doesn't mean that the RPCs
pushing state have finished - a pushing node may still be waiting for a response and so it
might increment its replication counter after the start of the test.
--
This message was sent by Atlassian JIRA
(v6.3.1#6329)