Uploaded image for project: 'Confluence Data Center'
  1. Confluence Data Center
  2. CONFSERVER-93299

Retention rule jobs spread to every node in the multi-node DataCenter cluster

      Issue Summary

      The Retention rule jobs, Version Removal, and Trash Removal jobs run on every node if they are scheduled and can't finish before the previous run is completed. This primarily affects 'Hard' jobs, which take longer than 'Soft' counterparts.

      Combined with the below bug, it either creates more load on the database or creates locks which makes threads wait for each other:

      This is reproducible on Data Center: yes

      Steps to Reproduce

      1. Create a multi-node DC cluster
      2. Disable all Retention jobs:
        • Version Removal (Soft)
        • Version Removal (Hard)
        • Trash Removal (Soft)
        • Trash Removal (Hard)
      3. Create a Space with lots of pages (+20k) deleted and staying in the Trash
      4. Configure the Retention rules and change 'Keep all' to 'Keep by deleted date' for Trash
      5. Enable 'Trash Removal (Hard)' and set the cronjob for running every minute

      Expected Results

      When the 'Trash Removal (Hard)' job gets triggered on one of the nodes, the following job execution attempts will be discarded with an 'Already running' message in the scheduler_run_details table.

      Actual Results

      If the following job execution attempt will be on the same node where the job was first triggered and is still running, it inserts a new record to the scheduler_run_details table with an 'Already running' message. If the following job execution will be on a different node, the job starts running on that node, too.

      Workaround

      Currently, the only workaround can be enabling the job, triggering it manually on one node, and disabling it again.

            [CONFSERVER-93299] Retention rule jobs spread to every node in the multi-node DataCenter cluster

            A fix for this issue is available in Confluence Data Center 9.1.0.
            Upgrade now or check out the Release Notes to see what other issues are resolved.

            Jordan Anslow added a comment - A fix for this issue is available in Confluence Data Center 9.1.0. Upgrade now or check out the Release Notes to see what other issues are resolved.

              5339cdd01cf4 Jeffery Xie
              9f7de485df51 Basar Beykoz
              Affected customers:
              4 This affects my team
              Watchers:
              11 Start watching this issue

                Created:
                Updated:
                Resolved: