Kafka 0.8.3 issue about afkak HOT 3 CLOSED

We don't currently test against 0.8.3—the last in the 0.8 series Afkak's integration tests run against is 0.8.2.2. This is controlled by KAFKA_ALL_VERS in the Makefile: https://github.com/ciena/afkak/blob/master/Makefile#L11

I'm not aware of any issue that happens after a half hour. That log message likely indicates that the Kafka broker is in a bad state. Usually we see this on resource-constrained systems: Kafka and ZooKeeper do not deal well with CPU or I/O starvation, and Kafka tends to get wedged in a bad state when anything disrupts its communications with ZooKeeper. As a first debugging step, does restarting Kafka restore it to an operational state?

from afkak.

I'm closing this issue as there isn't anything we can do about it without more information. If you have more to add, please feel free to reopen. Thanks for your interest in Afkak!

from afkak.

The multiple of 10 minutes (30 minutes) makes me think it's the idle connection drop, though it should reconnect automagically on the next message send.

from afkak.