Definitely no long pauses on the consumer. I see a minor collection every
second which uses up 0.1 or 0.2 seconds. That in itself seems a bit on the
higher side (~10-20% time spent in GC) but I don't think that would cause a
zk session timeout. Now getting gc stats on the zookeeper side is a bit
harder-- this is not a system we control!

So in your opinion, long gc pauses are the most likely explanation for this.

On Tue, Feb 5, 2013 at 8:27 PM, Jay Kreps <[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