-
Story
-
Resolution: Done
-
Normal
-
None
-
None
-
False
-
None
-
False
-
-
Similar to what we did for clustersync (see https://github.com/openshift/hive/pull/1093), split the machinepool controller out of hive-controllers and into its own statefulset.
Each replica controls a subset of MachinePools. May as well (factor out and) reuse the same algorithm as clustersync: modulo the UUID of the MP (or CD??) by the number of replicas and assign to that replica. Add a status field to the MP indicating which replica is handling it.