2014-10-27 - production
18:34	<manybubbles>	restarting elasticsearch servers to pick up new gc logging and to reset them into a "working" state so they can have their gc problem again and we can log it properly this time.	[production]