Our primary database provider rolled out an unscheduled update which affected our systems' ability to communicate at 1700 GMT. We contacted our database provider to roll back the change. No data was lost, but this did cause a backlog of tasks to build up, leading to system latency and app unavailability.