This incident has been resolved. The root cause was a misconfiguration performed by one of our infrastructure providers at 8:08 GMT-3. At 8:35 GMT-3 we had already migrated away 80% of the traffic, and then worked to reestablish service for the remaining 20%, which concluded at 10:38 GMT-3.
Update · UTC
We implemented a fix and are currently monitoring the result.
Monitoring · UTC
We are in contact with our CDN provider and expect a full resolution soon
While that, the migration to our CDN backup is in progress and working
Identified · UTC
One of our major CDN provider is experiencing issues
We're working on migrating the sites to a backup CDN