Due to a Meltdown and Spectre patch applied to our service node by our cloud provider, the backend services did not recover automatically as expected. This was unfortunately only discovered after an hour of outage. All services have subsequently been restored and are fully operational.