As seen on T231185 the table general_log_sampled is now growing around 1TB per week. It causes stalls on Tendril, we should create a cronjob or an event that either truncates it or deletes rows everyday.
Deleting rows is problematic with such a big table and how much writes its gets.
However, truncating it means we'll no longer have slow query reports...we need to find a balance of how many days we want be able to check on tendril.