Remove misleading ignore for sharded collections in configDB in move unsharded balancer policy

XMLWordPrintableJSON

    • Type: Task
    • Resolution: Unresolved
    • Priority: Major - P3
    • None
    • Affects Version/s: 8.3.0, 9.1.0-rc0, 9.0.0
    • Component/s: None
    • None
    • Catalog and Routing
    • CAR Team 2026-09-14
    • 🟩 Routing and Topology
    • None
    • None
    • None
    • None
    • None
    • None

      [Test only issue]

       

      In clusterHasShardedCollections, which is used to decide between moveChunk and moveCollection when the balancerShouldReturnRandomMigrations fail point is enabled, there is this code and comment:

          // Skip config.system.sessions if we are not draining as it isn't balanced as part of the random
    // migrations failpoint. If we are draining shards, though, we need to include this collection.
    if (!draining) {
        matchBuilder.append(CollectionType::kNssFieldName, BSON("$regex" << "^(?!config\\.).*"));
    }

       

       

      This isn't true anymore and since SERVER-105874, in random migrations failpoint mode we balance config.system.sessions.

      This ticket should remove that check. This will make the balancer respect the skipMoveCollectionThreshold setting  when only config.system.sessions can be balanced, rather than always preferring moveCollection.

            Assignee:
            Joan Bruguera Micó
            Reporter:
            Joan Bruguera Micó
            Votes:
            0 Vote for this issue
            Watchers:
            1 Start watching this issue

              Created:
              Updated: