In our kafka consumer logs, we're seeing the following messages:
2018-03-05 03:57:03,350 INFO
org.apache.kafka.clients.consumer.internals.AbstractCoordinator - Marking the
coordinator kafka08:9092 (id: 2147483639 rack: null) dead for group mygroup
2018-03-05 03:57:03,350 WARN
org.apache.kafka.clients.consumer.internals.ConsumerCoordinator - Auto-commit
of offsets {topic1-11=OffsetAndMetadata{offset=64888089, metadata=''}} failed
for group mygroup: Offset commit failed with a retriable exception. You should
retry committing offsets. The underlying error was: The request timed out.
After this message appears, the consumer no longer auto-commits offsets to the
broker, even though the timeout does not repeat (and the consumer continues to
consume messages from the topic partitions).
Is this expected behavior that once the coordinator is marked dead from a
single timeout, it no longer attempts to send auto-commit offsets? Shouldn't
the timeout apply only to a single attempt at
ConsumerCoordinator.sendOffsetCommitRequest(), and subsequent offset commits
would still go through?
We are using kafka client version 0.11.0.2 and kafka server version kafka-1.0.0.
