Between 09:32 and 11:05 CEST, the EVO-Cloud was
unavailable and requests returned a "502 Bad Gateway" error.

The cause was an expired security credential that the application uses
to access its own configuration store.

We renewed the credential and the service recovered at 11:05 CEST.
Monitoring confirmed availability from all regions. No data was lost or
affected at any point.

To prevent a recurrence we are adding advance alerting on credential
expiry and moving to a credential type that does not expire silently.