soarez opened a new pull request, #12174: URL: https://github.com/apache/kafka/pull/12174
When a controlled shutdown is requested the broker tries to communicate the state change to the controller via a heartbeat request. [1] In this test, the controller is not available so the request will fail. The current timeout behavior in a heartbeat request is to just keep retrying — which generally makes sense, just not in the context of a controlled shutdown. When a heartbeat request times out, if we are in the middle of a controlled shutdown, we shouldn't just retry forever but rather just give up on trying to contact the controller and proceed with the controlled shutdown. [1] https://github.com/apache/kafka/blob/f2d6282668a31b9a554563338f9178e2bba2833f/core/src/main/scala/kafka/server/BrokerLifecycleManager.scala#L217 *Summary of testing strategy* The test no longer fails ### Committer Checklist (excluded from commit message) - [ ] Verify design and implementation - [ ] Verify test coverage and CI build status - [ ] Verify documentation (including upgrade notes) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
