Since kafka is written in scala, when using a java client (producer or
consumer), do you pay a slight performance hit because you guys are
internally mapping back/forth to map a java type to a scala type?

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