heesung-sn opened a new issue, #23125: URL: https://github.com/apache/pulsar/issues/23125
### Search before asking - [X] I searched in the [issues](https://github.com/apache/pulsar/issues) and found nothing similar. ### Read release policy - [X] I understand that unsupported versions don't get bug fixes. I will attempt to reproduce the issue on a supported version of Pulsar client and Pulsar broker. ### Version pulsar-2.11 ### Minimal reproduce step It is hard to simulate this case, but essentially it seems like this could happen when the zk cluster is unstable or slow. ### What did you expect to see? Eventually, when the zk cluster regains stability, all brokers should see the same leader, and the leader should play its role such as load shedding. ### What did you see instead? The elected leader broker continued to play the follower role Indefinitely, and hence no load shedding logic occurred. Log that the elected broker thinks it's a follower. [metadata-store-10-1] INFO org.apache.pulsar.broker.PulsarService - This broker is a follower. Current leader is Optional[LeaderBroker(serviceUrl=http://xxx-broker-8.yyyy:8080/)] xxx-broker-8 ### Anything else? Potentially this is the root cause of this issue, https://github.com/apache/pulsar/issues/16145 ### Are you willing to submit a PR? - [X] I'm willing to submit a PR! -- 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]
