Increased error rates

8:21 PM PDT We are investigating increased provisioning latencies for new load balancers, delayed propagation of load balancer CloudWatch metrics, and inconsistent health check statuses from the DescribeTargetHealth API for Network Load Balancers in the US-EAST-1 Region.

10:08 PM PDT We can confirm delayed propagation of load balancer CloudWatch metrics and provisioning latencies for Network Load Balancers in the US-EAST-1 Region, and are continuing to work towards resolution. Connectivity to existing load balancers is not affected.

11:22 PM PDT Between 6:08 PM and 11:12 PM PDT, we experienced increased provisioning latencies for Network Load Balancers in the US-EAST-1 Region, which has now fully recovered. We are continuing to experience delays in the propagation of CloudWatch metrics for load balancers. The system used to generate these metrics has begun to recover, and we anticipate full recovery in the next few hours after we process the backlog of metrics.