Details
-
New Feature
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Won't Do
-
None
-
None
-
MXS-SPRINT-183, MXS-SPRINT-184
Description
Trigger demotion.sql if for some reason a down master comes back up and is replicating from the correct server. Right now since the monitor sees everything working correctly it doesn't execute rejoin and hence the demotion.sql is not triggered.
Here is and example of a sequence events where we would want the demotion.sql script to be triggered.
00:18:19 mariadb-0 MasterDown (this was due to networking issues where maxscale could not communicate with the node)
00:18:23 mariadb-1 NewMaster (mariadb-1 node promoted to Master)
00:18:24 mariadb-0 ServerUp (mariadb-0 original [failed] Master is accessible again without rebooting)
00:18:26 mariadb-0 NewSlave (mariadb-0 original [failed] Master rejoins as Slave)