Investigate performance regression when using shared buffers in RecordId

XMLWordPrintableJSON

    • Type: Task
    • Resolution: Gone away
    • Priority: Major - P3
    • None
    • Affects Version/s: None
    • Component/s: None
    • None
    • Execution Team 2022-09-19, Execution Team 2022-10-03
    • None
    • 3
    • None
    • None
    • None
    • None
    • None
    • None

      With the work done on SERVER-67471 we realised that swapping the Shared Buffer with a unique buffer manually managed by RecordId yielded significant performance improvements even if the branching itself wasn't taken.

      This task is about investigating why this is the case and to see if we can go back to the old behaviour while maintaining the performance characteristics.

        1. baseline.lookup.branchmiss.svg
          745 kB
          Jordi Olivares Provencio
        2. reproducer.patch
          4 kB
          Jordi Olivares Provencio
        3. shared-buffer.lookup.branchmiss.svg
          576 kB
          Jordi Olivares Provencio

            Assignee:
            Jordi Olivares Provencio
            Reporter:
            Jordi Olivares Provencio
            Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

              Created:
              Updated:
              Resolved: