Huhm, it seems like the coordinator is not re-queried when we get a NOT_COORD_FOR_GROUP error for OffsetCommitRequests, but it will be re-queried when we get the same error for HeartbeatRequests so the situation should fix itself within the heartbeat interval (by default within a couple of seconds).
Here it's about the not coord for group but the same principle of metadata state refresh applies to not_coord.
The consumer should reset itself without a need for a new instance.
This code:
https://github.com/zendesk/racecar/blob/master/lib/racecar/consumer_set.rb#L165
handles the
not_coordinator
by resetting the consumer.Based on this comment:
https://github.com/edenhill/librdkafka/issues/2630#issuecomment-598193001
Here it's about the not coord for group but the same principle of metadata state refresh applies to not_coord.
The consumer should reset itself without a need for a new instance.
Do you want me to fix this?