This incident has been resolved.

Server autoscaler halted and prevented new application servers from being created on demand.
Autoscaler has been restarted and services should be back to normal.