log.info("reading all : " + clustername); may be called every minute or more (depending on configuration) and prints this not terribly interesting message even if there is only one participant in the cluster and nothing whatsoever is happening. Would be better to log at debug level I think.