After recaching all activities and availabilities in our database, we have been monitoring our performance for a while and everything is back to normal.
We’ve learned a lot from this experience and we’re working on preventing these kind of issues in the future.