Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-31860

setFCV on a cluster can fail to call setFCV on shards if the config server primary's ShardRegistry is empty

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Closed
    • Priority: Major - P3
    • Resolution: Fixed
    • Affects Version/s: 3.6.0-rc2
    • Fix Version/s: 3.6.3, 3.7.1
    • Component/s: Sharding
    • Labels:
      None
    • Backwards Compatibility:
      Fully Compatible
    • Operating System:
      ALL
    • Backport Requested:
      v3.6
    • Sprint:
      Sharding 2017-11-13, Sharding 2017-12-04
    • Linked BF Score:
      0

      Description

      It's possible for the previous ShardRegistry reload to have failed (e.g., due to the majority snapshot not being available), and setFCV uses ShardRegistry::getAllShardIds(), which does not force a ShardRegistry reload.

      We could band-aid this by adding a ShardRegistry reload before the ShardRegistry::getAllShardIds(), or we could force a reload inside getAllShardIds() (and maybe also ShardRegistry::getNumShards(), a similar function that has the same issue).

        Attachments

          Activity

            People

            Assignee:
            jack.mulrow Jack Mulrow
            Reporter:
            esha.maharishi Esha Maharishi
            Participants:
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Dates

              Created:
              Updated:
              Resolved: