This outage was caused by an overheating event on two servers.
Around 2pm on Sunday, the 14th of August, two of our servers (TE-Server and HV1) powered off automatically to protect themselves from overheating.
We are still working on implementing additional methods of cooling for our systems, the redundancy systems were able to keep all customer-facing services online at reduced capacity.
Since starting the servers up again at 21:00, all system load has been redistributed across the two hosts and systems have returned to normal operating status.