Incident
Database server outage
ResolvedAt around 19:00 this evening a master database server behind the TwentyThree service failed, and as we were working to bring the service back online we identified a hardware malfunction (on a RAM block) on the server responsible for delivering the service. We quickly tried a quick recovery using the same hardware, but determined within a few minutes that the best viable option was to switch to using new hardware. The switch was made and all service restored at 19:44. During this interim the service was unavailable for a total of 25 minutes.
ResolvedThis incident has been resolved.
IdentifiedWe have identified an issue on a core database server that require the service to be switche to different hardware. This will cause emergency downtime over the next 10 minutes. This page will be kept up to date.