2013-11-02 - production
09:56	<apergos>	restarted twemproxy on ~10 of the api servers in eqiad in the lower memory/core group, was seeing a lot of entries in memcached-serious log and some processes churning on that	[production]