Introduce a failpoint prior to checkpoint completion

XMLWordPrintableJSON

    • Type: Task
    • Resolution: Unresolved
    • Priority: Critical - P2
    • None
    • Affects Version/s: None
    • Component/s: Checkpoints
    • None
    • Storage Engines - Persistence
    • SE Persistence backlog
    • 3

      Introduce a failpoint in __checkpoint_tree that triggers before checkpoint completion/checkpoint resolve, returning an error (without crashing the process). This will allow us to simulate checkpoint failures at this stage and verify that rollback logic executes correctly and the system remains consistent.

      Definition of Done

      • Introduce a failpoint that returns an error (does not crash) and exercises the checkpoint reset path (through __checkpoint_fail_reset)
      • Add testing to trigger this failpoint and validate behaviour.
      • Ensure verify does not report page discard or PALI inconsistencies after recovery.

        1. jira_comment.md
          6 kB
        2. full.diff
          155 kB
        3. failpoint+test.diff
          4 kB

            Assignee:
            [DO NOT USE] Backlog - Storage Engines Team
            Reporter:
            Mariam Mojid
            Votes:
            0 Vote for this issue
            Watchers:
            4 Start watching this issue

              Created:
              Updated: