Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-42234

Make txn_two_phase_commit_failover.js robust to all replica nodes running for election at the same time

    XMLWordPrintable

    Details

    • Backwards Compatibility:
      Fully Compatible
    • Operating System:
      ALL
    • Backport Requested:
      v4.2
    • Sprint:
      Sharding 2019-08-12, Sharding 2019-08-26, Sharding 2019-09-09
    • Linked BF Score:
      9

      Description

      The test repeatedly induces failovers on the coordinator shard. Currently, it's possible for all three nodes to run for election and therefore refuse to vote for anyone but themselves, and mongos times out after waiting for 20 seconds for the coordinator shard to have a primary.

      One idea is to only make two of the three replica nodes in the coordinator shard electable. The third replica node should then act as an "arbiter" in elections and vote for one of the other two.

        Attachments

          Issue Links

            Activity

              People

              • Votes:
                0 Vote for this issue
                Watchers:
                2 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: