It was broken apart because we checked if there is anything to commit in the first place (ie, do the check on one place)-- if we did not process any data, we don't need to commit.
Not sure why this check was removed (cf. my comment above). We can also rearrange the code as proposed if you think it's cleaner and add the check for both cases. [ Full content available at: https://github.com/apache/kafka/pull/5428 ] This message was relayed via gitbox.apache.org for [email protected]
