A moveRange with waitForDelete:true can succeed without running the range deletion after getting aborted by a DDL

XMLWordPrintableJSON

    • Type: Bug
    • Resolution: Unresolved
    • Priority: Major - P3
    • None
    • Affects Version/s: 7.0.0, 8.0.0, 8.2.0, 9.0.0-rc0, 8.3.0
    • Component/s: None
    • Cluster Scalability
    • ALL
    • Cluster Scalability Priorities
    • 200
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      The moveChunk/moveRange commands have a waitForDelete parameter, which can be set to true to wait for the range deletion to be executed on the donor.

      It's also possible that the migration is aborted by a DDL (e.g. createIndexes), in which case the migration will fail with a OrphanedRangeCleanUpFailed error code.

      However, after this error happens, we re-check the outcome of the migration and if we see that the chunk migration itself completed, then we will re-write the outcome to OK. As a consequence, we will return OK to the user despite the full command (move chunk and wait for orphan deletion) not being complete.

       

      A reproducer is attached.

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

              Created:
              Updated: