[ https://issues.apache.org/jira/browse/KAFKA-13917?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
MillieZhang updated KAFKA-13917: -------------------------------- Attachment: image-2022-10-21-15-53-14-689.png > Avoid calling lookupCoordinator() in tight loop > ----------------------------------------------- > > Key: KAFKA-13917 > URL: https://issues.apache.org/jira/browse/KAFKA-13917 > Project: Kafka > Issue Type: Improvement > Components: consumer > Affects Versions: 3.1.0, 3.1.1, 3.3.0, 3.1.2, 3.2.1 > Reporter: Viktor Somogyi-Vass > Assignee: Viktor Somogyi-Vass > Priority: Major > Fix For: 3.3.0, 3.2.1 > > Attachments: image-2022-10-21-15-52-20-402.png, > image-2022-10-21-15-53-14-689.png > > > Currently the heartbeat thread's lookupCoordinator() is called in a tight > loop if brokers crash and the consumer is left running. Besides that it > floods the logs on debug level, it increases CPU usage as well. > The fix is easy, just need to put a backoff call after coordinator lookup. > Reproduction: > # Start a few brokers > # Create a topic and produce to it > # Start consuming > # Stop all brokers > At this point lookupCoordinator() will be called in a tight loop. -- This message was sent by Atlassian Jira (v8.20.10#820010)