Well, it's hard to say... It depends on which part user prefers. If user does 
not care about the duplicates, he/she will expect the MM should complete the 
rebalance asap, then a quick evac from committing is expected.  Assume the 
network error is not transient and `default.api.timeout.ms` is set to MAX, the 
onPartitionsRevoked will never end. In this case keeping retrying offers little 
help. Right?

[ Full content available at: https://github.com/apache/kafka/pull/5492 ]
This message was relayed via gitbox.apache.org for [email protected]

Reply via email to