lianetm commented on PR #16749: URL: https://github.com/apache/kafka/pull/16749#issuecomment-2322419699
Actually, I would expect the behaviour that the documentation currently had: > If insufficient data is available the request will wait for that much data to accumulate and it would make sense to me if we wanted to extend that with something like : `...waiting for up to the`fetch.max.wait.ms` (and we link the 2 configs) Is that the intention? The current change is actually confusing, seems to suggest that it will wait for the "full" fetch.max.wait.ms even if the data becomes available? This is truly a broker-side behaviour I'm not fully familiar with (the consumer just passes the config to the broker), but looking at how the delayed fetch is handled on the broker it aligns with the expectation that it waits for the data to appear, for "at most" the max.wait (https://github.com/apache/kafka/blob/70dd577286de31ef20dc4f198e95f9b9e4479b47/core/src/main/scala/kafka/server/DelayedOperation.scala#L34-L36). -- 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
