We have rerouted traffic to work around the unavailability of some of our AWS EC2 instances and are currently seeing error rates going down. We will continue to monitor the situation and take action when necessary.
