Uploaded image for project: 'HBase'
  1. HBase
  2. HBASE-27720

TestClusterRestartFailover is flakey due to metrics assertion

    XMLWordPrintableJSON

Details

    • Task
    • Status: Resolved
    • Minor
    • Resolution: Fixed
    • 2.6.0, 3.0.0-alpha-3, 2.4.16, 2.5.4
    • 2.6.0, 3.0.0-alpha-4, 2.5.5, 2.4.18
    • test
    • None

    Description

      I'm seeing failures like this in PR,

      [ERROR] Failures: 
[ERROR] org.apache.hadoop.hbase.master.TestClusterRestartFailoverSplitWithoutZk.test
[ERROR]   Run 1: TestClusterRestartFailoverSplitWithoutZk>TestClusterRestartFailover.test:143 serverNode should be deleted after SCP finished expected null, but was:<d09295c85a7e,33451,1678875566216/SPLITTING/regionCount=0/lock=java.util.concurrent.locks.ReentrantReadWriteLock@1823f707[Write locks = 0, Read locks = 0]>
[ERROR]   Run 2: TestClusterRestartFailoverSplitWithoutZk>TestClusterRestartFailover.test:147 serverCrashSubmittedCount(8) should be equal expected:<4> but was:<8>
[ERROR]   Run 3: TestClusterRestartFailoverSplitWithoutZk>TestClusterRestartFailover.test:147 serverCrashSubmittedCount(12) should be equal expected:<4> but was:<12>

      Looks like subsequent runs would have passed, but for the firm metric count assertion.

      Attachments

        Issue Links

          links to

          Web Link GitHub Pull Request #5110

          Activity

            People

              ndimiduk Nick Dimiduk
              ndimiduk Nick Dimiduk
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: