2019-02-28 - production
10:00	<_joe_>	executing a rolling puppet run (2 server at a time per cluster, per dc) in eqiad,codfw as an HHVM restart will be triggered	[production]