Issue has been identified and resolved. Latency for all endpoints is now back to normal values.
We are investigating increased latency on some community endpoints.