Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-54818

Mitigate distributed lock acquisition unfairness

    XMLWordPrintable

    Details

    • Backwards Compatibility:
      Fully Compatible
    • Operating System:
      ALL
    • Linked BF Score:
      30

      Description

      After SERVER-53118 any operation waiting for a dist lock will wait for 20 seconds before returning a LockBusy error if couldn't acquire the lock. Additionally, notify all is being used when unlocking to wake up all waiting operations. If we have too many concurrent operations and they're taking too long, these two facts might end up starving a request. We should preserve the acquisition order using notify_one, and increase the timeout (for example to two hours).

        Attachments

          Issue Links

            Activity

              People

              Assignee:
              tommaso.tocci Tommaso Tocci
              Reporter:
              marcos.grillo Marcos José Grillo Ramirez
              Participants:
              Votes:
              0 Vote for this issue
              Watchers:
              4 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved:

                    PagerDuty