Seems not real failure, maybe this is caused by "collectors" data structure: hashset.



At 2012-03-06 19:07:22,"Bing Li" <sarah.libing@gmail.com> wrote:
Hi, all
I found that TestFailChainManager and TestLogicalConfigManager in Flume-0.9.4 failed with the following errors:


Running com.cloudera.flume.master.availability.TestFailChainManager
Tests run: 2, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 1.845 sec <<< FAILURE!


Results :


Failed tests: 
  testRandomAvailMan(com.cloudera.flume.master.availability.TestFailChainManager): expected:<..."%s\")}","collector [5","collector 1","collector 3]")> but was:<..."%s\")}","collector [3","collector 1","collector 5]")>


Tests run: 2, Failures: 1, Errors: 0, Skipped: 0




Running com.cloudera.flume.master.logical.TestLogicalConfigManager
Tests run: 15, Failures: 6, Errors: 0, Skipped: 0, Time elapsed: 26.17 sec <<< FAILURE!


Results :


Failed tests: 
  testFailoverLogicalTrans(com.cloudera.flume.master.logical.TestLogicalConfigManager): expected:<...8.svl.ibm.com", 3585[5 ) } ? < { lazyOpen => rpcSink( "svltest118.svl.ibm.com", 35854] ) } ? null > > >> but was:<...8.svl.ibm.com", 3585[4 ) } ? < { lazyOpen => rpcSink( "svltest118.svl.ibm.com",35855] ) } ? null > > >>
  testMultipleRefreshes(com.cloudera.flume.master.logical.TestLogicalConfigManager): expected:<...8.svl.ibm.com", 3585[5 ) } ? < { lazyOpen => rpcSink( "svltest118.svl.ibm.com", 35854] ) } ? null > > >> but was:<...8.svl.ibm.com", 3585[4 ) } ? < { lazyOpen => rpcSink( "svltest118.svl.ibm.com", 35855] ) } ? null > > >>
  testMultipleUpdates(com.cloudera.flume.master.logical.TestLogicalConfigManager): expected:<...8.svl.ibm.com", 3585[5 ) } ? < { lazyOpen => rpcSink( "svltest118.svl.ibm.com", 35854] ) } ? null > > >> but was:<...8.svl.ibm.com", 3585[4 ) } ? < { lazyOpen => rpcSink( "svltest118.svl.ibm.com", 35855] ) } ? null > > >>
  testNoPhysicalNode(com.cloudera.flume.master.logical.TestLogicalConfigManager): expected:<... "logicalSink( \"foo[3\" )" ) } ? < { lazyOpen => fail( "logicalSink( \"foo2]\" )" ) } ? null > >...> but was:<... "logicalSink( \"foo[2\" )" ) } ? < { lazyOpen => fail( "logicalSink( \"foo3]\" )" ) } ? null > >...>
  testPhysicalNodeLost(com.cloudera.flume.master.logical.TestLogicalConfigManager): expected:<... "logicalSink( \"foo[3\" )" ) } ? < { lazyOpen => fail( "logicalSink( \"foo2]\" )" ) } ? null > >...> but was:<... "logicalSink( \"foo[2\" )" ) } ? < { lazyOpen => fail( "logicalSink( \"foo3]\" )" ) } ? null > >...>
  testUpdateAfterUnmapAll(com.cloudera.flume.master.logical.TestLogicalConfigManager): expected:<...8.svl.ibm.com", 3585[5 ) } ? < { lazyOpen => rpcSink( "svltest118.svl.ibm.com", 35854] ) } ? null > > >> but was:<...8.svl.ibm.com", 3585[4 ) } ? < { lazyOpen => rpcSink( "svltest118.svl.ibm.com", 35855] ) } ? null > > >>


Tests run: 15, Failures: 6, Errors: 0, Skipped: 0




Are they REAL failures?