Last week, after extensive testing, we moved Object Storage to a new proxy setup. The infrastructure of the provider behind that setup now has an incident, which made part of the requests fail. We have switched back to the previous setup, and requests succeed again. We keep watching.
