Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-32122

Node killing could cause applier to fail on write rows event leading inconsistency

    XMLWordPrintable

Details

    • Bug
    • Status: Closed (View Workflow)
    • Critical
    • Resolution: Cannot Reproduce
    • 10.6
    • N/A
    • Galera
    • None

    Description

      3-node cluster with RQG load where node_2 is killed. Node_1 applier fails on deadlock error and drops from cluster, node_3 self leaves as it is alone.

      2023-09-07 11:50:07 2 [ERROR] Slave SQL: Could not execute Write_rows_v1 event on table test.table20_int_autoinc; Deadlock found when trying to get lock; try restarting transaction, Error_code: 1213; handler error HA_ERR_LOCK_DEADLOCK; the event's master log FIRST, end_log_pos 398, Internal MariaDB error code: 1213
      

      Attachments

        1. n1.cnf
          1 kB
        2. n2.cnf
          1 kB
        3. n3.cnf
          1 kB
        4. node1.err
          311 kB
        5. node2.err
          477 kB
        6. node3.err
          830 kB
        7. transactions.zz
          1.0 kB
        8. transactions-bkp_1.yy
          4 kB
        9. transactions-bkp.yy
          5 kB

        Issue Links

          Activity

            People

              janlindstrom Jan Lindström
              janlindstrom Jan Lindström
              Votes:
              1 Vote for this issue
              Watchers:
              5 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:

                Git Integration

                  Error rendering 'com.xiplink.jira.git.jira_git_plugin:git-issue-webpanel'. Please contact your Jira administrators.