At 07:21 our cloud hosting provider reported that a memory DIMM of the underlying infrastructure experienced a failure, causing a hypervisor servicing our instances to enter an unhealthy state. And a migration was performed to move our instances onto healthy infrastructure.