Details
-
Task
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Done
-
None
Description
For multi master loop , we do provision slaves before introducing them in the loop ,
replacing a dead node happens twice a year in average , in this case if the new slave have same domain id and server id for the one that it replace , it will simply cancel the events originating from himself after the backup , so we do use other server_id, domin_id for every new node.
This lead to unnecessary big GTID after a while that we would like to to shrink
We think that similar syntax would be useful :
RESET MASTER MASTER_SERVER = ID
RESET MASTER MASTER_DOMAIN = ID