Production server hardware failure
Incident Report for Electric Imp
Resolved
Underlying hardware failed at 02:42 UTC, devices automatically migrated to other servers in the cluster by around 02:45 UTC.

A replacement server was added to the cluster at 02:55 UTC, and agents for offline devices began to restart.

Agents on the affected server would have seen restarts.
Posted Apr 18, 2017 - 03:11 UTC