Uploaded image for project: 'MariaDB ColumnStore'
  1. MariaDB ColumnStore
  2. MCOL-945

MariaDBReplication slave messages is incorrectly sending updates to PM nodes

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Closed (View Workflow)
    • Priority: Major
    • Resolution: Fixed
    • Affects Version/s: 1.0.11
    • Fix Version/s: 1.0.12, 1.1.1
    • Component/s: None
    • Labels:
      None
    • Sprint:
      2017-21

      Description

      This issue reported by a customer, they dont have the Local Query Feature set, but it shows that replication messages are being sent to the pm nodes. Should only be the UM.

      I am concerned about this in the debug.log from the startup:
      Sep 28 13:48:26 mcs1-um1 ProcessMonitor[33788]: 26.625489 |0|0|0| I 18 CAL0000: MSG RECEIVED: Run Master Replication script
      Sep 28 13:48:26 mcs1-um1 ProcessMonitor[33788]: 26.625545 |0|0|0| D 18 CAL0000: changeMyCnf function called for master
      Sep 28 13:48:26 mcs1-um1 DMLProc[42458]: 26.631175 |0|0|0| I 20 CAL0002: DMLProc will rollback 0 tables.
      Sep 28 13:48:26 mcs1-um1 DMLProc[42458]: 26.641852 |0|0|0| I 20 CAL0002: DMLProc finished rollbackAll.
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.720853 |0|0|0| D 18 CAL0000: changeMyCnf function successfully completed
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.720937 |0|0|0| D 18 CAL0000: runMasterRep function called
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.721950 |0|0|0| D 18 CAL0000: cmd = /usr/local/mariadb/columnstore/bin/master-rep-columnstore.sh --password='RMSAdmin' --installdir=/usr/local/mariadb/columnstore --hostIP=10.232.52.87 > /tmp/master-rep-columnstore-um1.log 2>&1
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.771243 |0|0|0| D 18 CAL0000: master-rep-columnstore.sh: Successful return for node um1
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.771297 |0|0|0| D 18 CAL0000: cmd = /usr/local/mariadb/columnstore/bin/master-rep-columnstore.sh --password='RMSAdmin' --installdir=/usr/local/mariadb/columnstore --hostIP=10.232.52.160 > /tmp/master-rep-columnstore-um2.log 2>&1
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.819111 |0|0|0| D 18 CAL0000: master-rep-columnstore.sh: Successful return for node um2
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.819164 |0|0|0| D 18 CAL0000: cmd = /usr/local/mariadb/columnstore/bin/master-rep-columnstore.sh --password='RMSAdmin' --installdir=/usr/local/mariadb/columnstore --hostIP=10.232.52.137 > /tmp/master-rep-columnstore-um3.log 2>&1
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.865990 |0|0|0| D 18 CAL0000: master-rep-columnstore.sh: Successful return for node um3
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.866040 |0|0|0| D 18 CAL0000: cmd = /usr/local/mariadb/columnstore/bin/master-rep-columnstore.sh --password='RMSAdmin' --installdir=/usr/local/mariadb/columnstore --hostIP=10.232.52.80 > /tmp/master-rep-columnstore-pm1.log 2>&1
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.913374 |0|0|0| D 18 CAL0000: master-rep-columnstore.sh: Successful return for node pm1
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.913426 |0|0|0| D 18 CAL0000: cmd = /usr/local/mariadb/columnstore/bin/master-rep-columnstore.sh --password='RMSAdmin' --installdir=/usr/local/mariadb/columnstore --hostIP=10.232.52.154 > /tmp/master-rep-columnstore-pm2.log 2>&1
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.960529 |0|0|0| D 18 CAL0000: master-rep-columnstore.sh: Successful return for node pm2
      Sep 28 13:48:35 mcs1-um1 ProcessMonitor[33788]: 35.960580 |0|0|0| D 18 CAL0000: cmd = /usr/local/mariadb/columnstore/bin/master-rep-columnstore.sh --password='RMSAdmin' --installdir=/usr/local/mariadb/columnstore --hostIP=10.232.52.61 > /tmp/master-rep-columnstore-pm3.log 2>&1
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.013787 |0|0|0| D 18 CAL0000: master-rep-columnstore.sh: Successful return for node pm3
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.013838 |0|0|0| D 18 CAL0000: cmd = /usr/local/mariadb/columnstore/bin/master-rep-columnstore.sh --password='RMSAdmin' --installdir=/usr/local/mariadb/columnstore --hostIP=10.232.52.142 > /tmp/master-rep-columnstore-pm4.log 2>&1
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.064471 |0|0|0| D 18 CAL0000: master-rep-columnstore.sh: Successful return for node pm4
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.064527 |0|0|0| D 18 CAL0000: cmd = /usr/local/mariadb/columnstore/bin/master-rep-columnstore.sh --password='RMSAdmin' --installdir=/usr/local/mariadb/columnstore --hostIP=10.232.52.148 > /tmp/master-rep-columnstore-pm5.log 2>&1
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.111411 |0|0|0| D 18 CAL0000: master-rep-columnstore.sh: Successful return for node pm5
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.111493 |0|0|0| D 18 CAL0000: runMasterRep: masterlogfile=mysql-bin.000006, masterlogpos=3425
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.111542 |0|0|0| D 18 CAL0000: runDisableRep function called
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.111634 |0|0|0| D 18 CAL0000: cmd = /usr/local/mariadb/columnstore/bin/disable-rep-columnstore.sh --password='RMSAdmin' --installdir=/usr/local/mariadb/columnstore > /tmp/disable-rep-columnstore.log 2>&1
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.139146 |0|0|0| D 18 CAL0000: disable-rep-columnstore.sh: Successful return
      Sep 28 13:48:36 mcs1-um1 ProcessMonitor[33788]: 36.139221 |0|0|0| I 18 CAL0000: MASTERREP: ACK back to ProcMgr return status = 0
      Sep 28 13:48:59 mcs1-um1 ProcessMonitor[33788]: 59.552185 |0|0|0| I 18 CAL0000: MSG RECEIVED: Update Calpont Config file
      Sep 28 13:48:59 mcs1-um1 ProcessMonitor[33788]: 59.552506 |0|0|0| I 18 CAL0000: UPDATECONFIGFILE: Completed

      Show processlist on um1 shows the expected replication running to um2 and um3.
      The log output above shows a successful response from each of the pm machines.
      We disabled the local pm query option last week.
      Indeed there is not a mysqld process running on any of the pm's
      How could it get a successful response from a machine that is not running mysqld?
      Why did it even try to start replication on the pm's?

        Attachments

          Activity

            People

            Assignee:
            dleeyh Daniel Lee
            Reporter:
            hill David Hill (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.