I just saw all my test units restart, agents and devices. Nothing I did seems to have triggered it. Was something scheduled?
All occurring around 2015-04-14 21:15 UTC
The developer server’s agents did an unscheduled restart, yes. An issue which was fixed last month - but not yet deployed - cropped up again. Didn’t affect production servers, and the restart picked up the patch so it shouldn’t happen again.
One of the many reasons we’re looking forward to having Erlang running the agents which can be live patched without downtime…