An upstream provider suffered issues, causing the cluster to failover.
During this he cluster suffered an issue whilst migrating services, causing several the restoration of some sites to fail.
The entire web cluster had to be restarted to fix the issue.
All nodes are now functioning as normal.