Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-48046

{replSetFreeze: 0} done in RollbackTest.restartNode() should be resilient of network errors

    • Type: Icon: Task Task
    • Resolution: Fixed
    • Priority: Icon: Major - P3 Major - P3
    • 4.2.7, 4.4.0-rc6, 4.7.0
    • Affects Version/s: None
    • Component/s: Replication
    • Labels:
      None
    • Fully Compatible
    • v4.4, v4.2
    • 19

      In RollbackTest.restartNode we try to freeze and unfreeze the current secondary to prevent elections from happening when we restart the primary. We use ReplSetTest.freeze to freeze which handles network errors (see SERVER-43237). But when we try to unfreeze, we use a vanilla adminCommand which does NOT handle network errors. However, if the current secondary is in rollback, the node would close all user connections so the this replSetFreeze command could fail with network errors for the same reasons as described in SERVER-43237. So I think we should also make the "unfreeze" part in RollbackTest.restartNode() resilient of network errors.

            Assignee:
            lingzhi.deng@mongodb.com Lingzhi Deng
            Reporter:
            lingzhi.deng@mongodb.com Lingzhi Deng
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Created:
              Updated:
              Resolved: