It seems many endpoints end up in 500 errors. We are investigating.
The error seems to originate in a connection error between the application servers and the databases.
It seems other customers of our hosting provider complain that hostnames for the databases cannot be resolved.
We have opened a support ticket, OVHCloud status page does not reflect the incident yet.
UPDATE: A similar incident has been reported on OVHCloud's status page: https://public-cloud.status-ovhcloud.com/incidents/9qmxv0gbky43
UPDATE: OVH announces that the incident should be fixed, but are still monitoring.
UPDATE: Our internal monitoring shows that errors are reduced a lot, but not down to zero yet.
UPDATE: Our internal monitoring does not show any errors anymore.