Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-31446

_getNextSessionMods can get stuck in a write conflict retry loop

    • Type: Icon: Bug Bug
    • Resolution: Fixed
    • Priority: Icon: Major - P3 Major - P3
    • 3.6.0-rc3
    • Affects Version/s: None
    • Component/s: Sharding
    • Labels:
      None
    • Fully Compatible
    • ALL
    • Sharding 2017-10-23, Sharding 2017-11-13
    • 0

      1. Shard receives _getNextSessionMods command.
      2. Command tries to query oplog and hits a WriteConflictException
      3. QueryYield::yieldAllLocks is called, but it returns early because it's using DBDirectClient so it does not throw away the snapshot (Note: First global lock is taken by AutoGetActiveCloner).
      4. Query loop retries, but since it never resets the state during yield, it gets a WCE again, so step#2 is repeated.

            Assignee:
            randolph@mongodb.com Randolph Tan
            Reporter:
            randolph@mongodb.com Randolph Tan
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Created:
              Updated:
              Resolved: