Hi Jagadish,

This is a rather late reply but I don't think I understand of the effect of
changing job.container.thread.pool.size config in the synchronous case very
well.
Suppose I have an input topic partitioned by memberID processed by a
synchronous job.
Does increasing container thread pool size still guarantee that the
messages will be processed serially for messages with the same key?
For example, if I had 2 messages msg1 followed by msg2 with the same key
then is it guaranteed that msg1 will finish processing before starting to
process msg2?

My actual situation is that we currently have a job that has maxed out the
allowed number of containers but is still not processing messages fast
enough.
We want to speedup processing but we would like to avoid increasing the
number of partitions of the input topic.
The job is IO bound and has the requirement that messages with the same key
are processed serially.

Thanks,
Xiaochuan Yu

On Thu, Aug 10, 2017 at 11:28 PM Jagadish Venkatraman <
[EMAIL PROTECTED]> wrote:
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB