Uploaded image for project: 'MariaDB MaxScale'
  1. MariaDB MaxScale
  2. MXS-5353

MaxScale 24.02.3 fatal signal 6 after mariadb master is down

    XMLWordPrintable

Details

    • Bug
    • Status: Needs Feedback (View Workflow)
    • Major
    • Resolution: Unresolved
    • 24.02.3
    • None
    • None
    • OS = Linux maxscale-02 4.18.0-408.el8.x86_64 #1 SMP Mon Jul 18 17:42:52 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux
      MaxScale = 24.02.3
      MariaDB (master) = 10.11.7
      mariadbd --print-defaults = attach mariadb.defaults

    Description

      We have three maxscales with the same configuration, to which the traffic is distributed.
      Configuration synchronization is configured on all instances.
      After the fall of the mariadb master, the first maxscale-1 stopped responding to maxctrl requests, such as maxctrl show maxscale.
      But a new master was chosen and the slaves were switched.
      After restarting the service maxscale-1, it started responding to requests, but the other two maxscales disappeared from the synchronization status. Maxscale-2 and -3 continued to not respond to requests and constantly crash.
      After that I deleted the local configuration file /var/lib/maxscale/maxscale-config.json (in attach maxscale-config-1.json) on maxscale-2 server, deleted out the parameters related to the configuration synchronization in the main file /etc/maxscale.cnf. Only after that maxscale-2 started. Log maxscale_2.log from maxscale-2.

      Attachments

        1. maxscale-config-1.json
          37 kB
        2. maxscale_2.log
          2.54 MB
        3. mariadb.defaults
          3 kB

        Activity

          People

            Unassigned Unassigned
            dabbler Yuriy
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

            Dates

              Created:
              Updated:

              Git Integration

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