Uploaded image for project: 'Confluence Data Center'
  1. Confluence Data Center
  2. CONFSERVER-17201

Unclustered instance stopps with cluster panic during db backup since Confluence V3


    • Icon: Bug Bug
    • Resolution: Duplicate
    • Icon: High High
    • None
    • 3.0.1
    • None
    • Tomcat 6.0.18, Java 1.5.0_17, jdbc-mysql 5.0.8

      Without changing any other things in our setup, after Updating Confluence from 2.7 to 3.0.1 the custer savety manager stops the application with "cluster panic" neary every night at database backup time (7:50am).

      • It never stops at other times (, it's no "other running instance" problem or something like this).
      • This never happens during database backups with Confluence before 3.x

      I've attached a log with ClusterSavetyManager's loglevel set at DEBUG. Because it's a company-internal wiki, Confluence is totally idle at backup time. You'll see the successful checks just before the db backup, the shutdown and the successful checks afterward.

      Notice, that the check @195030.054 (Worker-3) seems to be stuck in waiting for the database and the scheduled check @ 195100.026 (Worker-9) is "passing by". In the other hand, Worker-3 finishes the check successful afterwards.

      From this i would derive, that no database connection timeout have happended but just a short "freeze" caused by the database dump operations.

      IMHO there should just one check running at a time. Did some kind of locking become unfunctional?

            Unassigned Unassigned
            9ebf83f4b609 Guido Jäkel
            0 Vote for this issue
            1 Start watching this issue