This incident has been resolved.

A fix has been implemented and we are monitoring the results.

A sudden spike in traffic overwhelmed our servers and even the autoscaling. We're looking into improving the autoscaler as systems seem to be getting back up.