Optimise duplicate check algorithm for timestamp safe unique index

XMLWordPrintableJSON

    • Type: Improvement
    • Resolution: Won't Fix
    • Priority: Major - P3
    • None
    • Affects Version/s: None
    • Component/s: Storage
    • Storage Engines 2019-05-20, Storage Engines 2019-06-03
    • 0
    • 8
    • None
    • 3
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      Timestamp safe unique index for 4.2 incorporates a duplicate key check/prevention algorithm. Older Unique index used to prevent duplicate keys in a different way. We put this algorithm to be timestamp safe.

      Linkbench (LOAD_NODE_BULK) shows a performance penalty due to this check. This ticket tracks the effort in optimising the algorithm and get the performance back.

        1. batch_ins.py
          1 kB
          Sulabh Mahajan
        2. linkbench_simu.py
          4 kB
          Sulabh Mahajan
        3. uniq_idx.py
          4 kB
          Sulabh Mahajan
        4. uniq_idx.sh
          2 kB
          Sulabh Mahajan

            Assignee:
            Sulabh Mahajan
            Reporter:
            Sulabh Mahajan
            Votes:
            0 Vote for this issue
            Watchers:
            10 Start watching this issue

              Created:
              Updated:
              Resolved: