Replica fails to restart if the status is checked before it's warm

XMLWordPrintableJSON

    • Type: Bug
    • Resolution: Duplicate
    • Priority: Major - P3
    • None
    • Affects Version/s: 4.2.16
    • Component/s: None
    • None
    • ALL
    • Hide

      Remove sleeps in the backport BACKPORT-10473. This never fails when run locally, but has a good chance to fail with burn-in tests at waterfall. Apparently the race requires the server to be resource constrained to show up.

      Show
      Remove sleeps in the backport BACKPORT-10473. This never fails when run locally, but has a good chance to fail with burn-in tests at waterfall. Apparently the race requires the server to be resource constrained to show up.
    • Sharding 2021-11-15
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      There is some race that makes the test flaky. Symptoms: a test kills and restarts replica fast, after restart it checks the status immediately. The replica never becomes either primary or secondary.

            Assignee:
            Andrew Shuvalov (Inactive)
            Reporter:
            Andrew Shuvalov (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

              Created:
              Updated:
              Resolved: