Github user edvorkin commented on the issue:
https://github.com/apache/zookeeper/pull/150
+1 one here. This feature is necessary for running Zookeeper in the cloud
under AWS ASG. Every time node fails, ASG reassigns new IP for new zookeeper
and there is no way kafka will know about it. We need to treat zookeeper
servers as cattle, not pets, and kill and spin new one at will without
affecting kafka. ---
