Comment by inglor

They addressed the issue by scaling up but the scale up itself ended up lagging behind the database. If that attack happens again presumably the bigger database + caching would mitigate it.