Inspect CAR test hooks for bugs where we misadvertently skip or ignore a validation

    • Type: Task
    • Resolution: Unresolved
    • Priority: Major - P3
    • None
    • Affects Version/s: None
    • Component/s: None
    • Catalog and Routing
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      We have recently detected multiple issues where the validations on our test hooks are broken or its results are ignored, causing Server bugs to slip through, e.g.:

      • SERVER-130586 The CheckMetadataConsistencyInBackground hook swallows failures in jstests passthrough suites
      • SERVER-129905 The check_uuids_consistent_across_cluster hook always succeeds without any checks due to broken config.chunks aggregation  
      • SERVER-131740 Exception in ContinuousAddRemoveShard thread loop does not trigger test failure
      • SERVER-134524 The ContinuousAddRemoveShard validation the shard is empty on decommission always succeeds due to broken iteration code

       

      This investigation ticket should:

      • Analyze existing test hooks to discard more bugs similar to the ones above.
      • Investigate how to prevent those bugs going forward (e.g. via static analyzers, adding tests for the hooks, AI analysis, etc.) - any followup work can be done in a separate SERVER ticket.

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

              Created:
              Updated: