Uploaded image for project: 'Compass '
  1. Compass
  2. COMPASS-6998

Investigate changes in SERVER-76001: Add serverStatus metrics for PeriodicThreadToAbortExpiredTransactions

    • Type: Icon: Investigation Investigation
    • Resolution: Done
    • Priority: Icon: Major - P3 Major - P3
    • No version
    • Affects Version/s: None
    • Component/s: None
    • None
    • Not Needed

      Original Downstream Change Summary

      New serverStatus field:
      "metrics.abortExpiredTransactions.passes"

      A 'pass' is scheduled periodically, and aborts transactions older than 'transactionLifetimeLimitSeconds'. "passes" is incremented after a successful pass.

      For Triage: If the metric stops incrementing, the thread could be stuck trying to abort a transaction.

      Description of Linked Ticket

      This thread is imperative to ensure forward progress of the system. Since we've seen multiple hangs where the thread gets stuck, we should add serverStatus metrics which track the number of of transactions killed by the thread.

      Additionally, we should consider adding a log line to indicate which session(s) the thread is trying to abort.

            Assignee:
            Unassigned Unassigned
            Reporter:
            backlog-server-pm Backlog - Core Eng Program Management Team
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Created:
              Updated:
              Resolved: