GitHub user hiloboy0119 opened a pull request:
https://github.com/apache/kafka/pull/7
Modified the async producer so it re-queues failed batches.
I'm working on an application that needs the throughput offered by an async
producer but also needs to handle send failures gracefully like a sync
producer. I modified the ProducerSendThread so it will re-queue failed
batches. This allowed me to determine the behavior I wanted from my producer
with the queue config parameters.
A "queue.enqueue.timeout.ms=0" allowed me to get runtime exceptions when
sends failed often enough to fill the queue. This also allowed me to use
"queue.buffering.max.messages" to control how tolerant the application is to
network blips.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/gradientx/kafka async_requeue
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/kafka/pull/7.patch
----
----
