The test failed randomly during last Infinispan testsuite run on Rhel5_x86, OpenJDK6.

You can find the log here:

http://jenkins.mw.lab.eng.bos.redhat.com/hudson/job/edg-60-ispn-testsuite-rhel/jdk=openjdk-1.6.0-local,label=RHEL5_x86/27/testReport/org.infinispan.lock.singlelock.replicated.pessimistic/InitiatorCrashPessimisticReplTest/testInitiatorNodeCrashesBeforeCommit/

The bug doesn't appear when the test is run separately.