[KAFKA-17506] KRaftMigrationDriver initialization race condition

• Type: Bug
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 3.8.0, 3.7.1, 3.9.0
• Fix Version/s: 3.9.0, 3.7.2, 3.8.1
• Assignee: David Arthur
• Reporter: David Arthur
• Created: 10/Sep/24 00:17
• Updated: 11/Sep/24 15:12
• Resolved: 11/Sep/24 15:12

Description

There is a race condition between KRaftMigrationDriver running its first poll() and being notified by Raft about a leader change. If onControllerChange is called before RecoverMigrationStateFromZKEvent is run, we will end up getting stuck in the INACTIVE state.

This is likely a very narrow race condition that is not likely to be seen in practice. It was found when debugging a flaky integration test.

If encountered, this can be worked around by restarting the stuck controller. Leadership will move to another controller which will allow the migration to be started.

Issue Links

• links to: GitHub Pull Request #17147