On of the API workers was down, which resulted in failure for a portion of the requests.
IP was redirected to a functional worker while we are investigating.
EDIT: Fixed