The incident is now resolved, we are sorry about the impact, and we are available to discuss the details if needed.
Monitoring · UTC
We implemented a fix and have re-scheduled most of the impacted failing service.
The initial incident triggering 404 was handled very shortly after it was detected, we noticed that some users were still experiencing some 503 due to an automatic garbage collection of some network resources.
Then we implemented a fix to recover the data that were garbage collected.
Identified · UTC
The recovery was done under 5m for most app, but a portion of user were still facing 503, we are currently trying to resolve this issue.
Investigating · UTC
We are currently investigating this incident, it seems it was caused by a configuration error from our side.