Page History

...

Producer and admin client use a default retries config value of Integer.MAX_VALUE and already rely on time based timeouts be default already (cf KIP-91 and KIP-533). They would still respected the deprecated retries config but log a warning if used. For Kafka Streams the retires config would be ignored (we only keep it to not break code that might set it) and log a warning if used. It The default retries value in Kafka Streams is 0 and we want to have a more robust default configuration. Note that the default retries values of 0 does not apply the embedded producer or admin client. Only if the user explicitly sets retries the embedded producer and admin client configs would we changed (this KIP does not change this behavior).

Furthermore, we propose to catch all client TimeoutException in Kafka Streams instead of treating them as fatal, and thus to not rely on the consumer/producer/admin client to handle all such errors. If a TimeoutException occurs, we skip the current task and move to the next task for processing (we will also log a WARNING for this case to give people inside which client call did produce the timeout exception). The failed task would automatically be retired in the next processing loop. Because other tasks a are processed until a task is retried, we don't have to worry about a busy wait situation. Even if a thread would have only a single task, the clients internal exponential retries would avoid busy waiting.

...

Last, the admin client is used within the group leader to collect topic metadata and to create internal topics if necessary. If those calls fails, they are retried within Kafka Streams re-using the admin client's retries config. Because admin retries will be deprecated, we should not re-use it any longer for this purpose. The current retry loop is across multiple admin client calls that are issues interleaved. This interleaved retry logic should be preserved. Furthermore, we should not retry infinitely to avoid that the leader is stuck forever (even if it would be removed from the group by the group coordinator after a timeout anyway that is set to max.poll.interval.ms). To avoid to drop dropping out of the consumer group, the retry loop should be stopped before we hit the timeout. We propose to use a 50% threshold, i.e., half of max.poll.interval.ms.

...