15% of services deployed in Frankfurt are unavailable due to a network failure of the underlying hypervisors.
Hypervisor failures happen and services should be automatically migrated - in this case, they didn’t automatically recover.
Our engineering team is investigating.