At 11:49 CEST an Operating System error on one of our databases caused the storage zone to abruptly shut down. The system recovered itself immediately, but the sudden shutdown left some files in inconsistent states.
At 11:54 CEST most stores in the affected storage zone was back online.
To preserve data integrity the remaining affected stores were put in maintenance mode while our automated tools went over every table to ensure they could be safely recovered. At 12:34 the last store was back in full production.
We will keep investigating the operating system error and will take appropriate measures to ensure increased reliability and data integrity.