Uploaded image for project: 'MariaDB MaxScale'
  1. MariaDB MaxScale
  2. MXS-978

Support for stale master in case of restart

    XMLWordPrintable

    Details

    • Type: New Feature
    • Status: Closed (View Workflow)
    • Priority: Major
    • Resolution: Fixed
    • Affects Version/s: 2.0.1
    • Fix Version/s: 2.0.2
    • Component/s: maxadmin
    • Labels:
      None

      Description

      In case where replication monitor gets a stale master status (slaves are down) and maxscale gets restarted, master loses the stale master status and no writes can happen.

      Suggest to track this or at least open the possibility to set stale master manually (not possible through maxadmin yet)

      Scenario
      A is master and die
      B and C was valide slave

      B is elected as new master

      A is stopped and get some unknow sate

      Remaining slave C

      • Can be , stopped
      • Can still be attached to A
      • Can be reset replication

      In all scenarios an external arbitrator forcing B as master , should enable routing to B via maxscale

      We propose to introduce a new state masterforced
      masterforced wins the master status if it conflict with mastermonitored
      when both state become equal then maxscale can re trust the monitoring information and reset masterforced with a simple masterflag

        Attachments

          Issue Links

            Activity

              People

              Assignee:
              markus makela markus makela
              Reporter:
              tanj Guillaume Lefranc
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved:

                  Git Integration