cmccabe commented on PR #13390: URL: https://github.com/apache/kafka/pull/13390#issuecomment-1468497417
cc @ijuma @mumrah @hachikuji So about the thread names thing. I’m open to changing the thread names to “kebab case” (i.e. `my-thread-name`) I do think in the context of JUnit we definitely need to have `broker-0-my-thread-name` and `controller-0-my-thread-name`. I find myself looking at JUnit backtraces all too often, and having 6 distinct threads named `my-thread-name` just doesn't work for me. So then the big question becomes whether we would want the prefixes in prod or not. The "pro" case is that it simplifies the code to just unconditionally do that, and avoid cases where someone accidentally forgets to set the prefix. The con case is that we should know what node we’re on, so the information is redundant. Although I’ve seen people do weird things like combine several process backtraces into one file or send ZK and Kafka logs all to the same file. So I don’“t truly believe the “we’ll never need it” case. Maybe “we rarely need it” or “we won’t need it if people are reasonable” -- 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: jira-unsubscr...@kafka.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
