Comment 1 for bug 1373513

I wonder if this is a result of something happening in terms of overall system load and perhaps flushing to disk. Experimenting with some ideas like changing the IO scheduler might be worth a shot.

Also Iv'e been considering a proposal to attach a Cinder Volume to our OpenStack instance and use that as our VG in gating rather than the loopback file. Don't have any data to say this is better or worse, but an idea I've kicked around for a bit now.