Uploaded image for project: 'MariaDB MaxScale'
  1. MariaDB MaxScale
  2. MXS-1594

MASTER_PASSWORD not being correctly used when triggering a switchover

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Closed (View Workflow)
    • Priority: Blocker
    • Resolution: Fixed
    • Affects Version/s: 2.2.1
    • Fix Version/s: 2.2.1
    • Component/s: mariadbmon
    • Labels:
      None
    • Environment:
      CentOS 7

      Description

      Folks,

      I found another problem guys when you do the switchover, you need to enter the CHANGE MASTER TO on the new slave to replicate off the new master. And then, a new file `master.info` is created with the information for the replication threads to start the replication. A couple of information is username and password, which in this case are equivalent to the replication_user and replication_password you set under the mysqlmon configurations for the Maxscale. Look at this:

      #: when Maxscale does the switchover, the master.info looks like...
      [root@box02 mysql]# cat master.info
      33
       
      4
      192.168.50.11
      repl
      'KP
      3306
      60
      0
       
      #: when I enter the change master to manually to investigate why the replication wasn't OK when set by the switchover process
      [root@box02 mysql]# cat master.info
      33
       
      4
      192.168.50.11
      repl
      123456
      3306
      60
      0
      

      You can see that the password for the user `repl` isn’t `'KP` but `123456`…maybe something wrong going on with using the var after reading the replication_password under the monitor configuration?

      After the switchover done by Maxscale, I’m always getting:

      Last_IO_Errno: 1045
      Last_IO_Error: error connecting to master 'repl@192.168.50.12:3306' - retry-time: 60  maximum-retries: 86400  message: Access denied for user 'repl'@'192.168.50.11' (using password: YES)
      

      My monitor configurations

      [root@maxscale maxscale.cnf.d]# cat cluster-monitor.cnf
      [cluster-monitor]
      type=monitor
      module=mysqlmon
      user=maxuser
      password=ACEEF153D52F8391E3218F9F2B259EAD
      monitor_interval=10000
      backend_connect_timeout=3
      backend_write_timeout=2
      backend_read_timeout=1
      backend_connect_attempts=1
      journal_max_age=28800
      script_timeout=90
      servers=prod_mariadb01,prod_mariadb02
      auto_failover=1
      failover_timeout=1
      master_failure_timeout=2
      replication_user=repl    #: <<<
      replication_password=123456  #: <<<
      detect_replication_lag=false
      detect_stale_master=true
      detect_stale_slave=true
      mysql51_replication=false
      multimaster=false
      detect_standalone_master=true
      failcount=5
      allow_cluster_recovery=true
      ignore_external_masters=false
      events=master_down,master_up,slave_down,slave_up,server_down,server_up,synced_down,synced_up,donor_down,donor_up,ndb_down,ndb_up,lost_master,lost_slave,lost_synced,lost_donor,lost_ndb,new_master,new_slave,new_synced,new_donor,new_ndb
      switchover_timeout=90
      verify_master_failure=true
      auto_rejoin=false
      

      Can you folks have a look at this issue?

        Attachments

          Activity

            People

            Assignee:
            esa.korhonen Esa Korhonen
            Reporter:
            wagnerbianchi Wagner Bianchi (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            2 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.