Between 7:16pm and 7:19pm UTC, we observed an elevated error rate and higher-than-normal latency in our US region for our Ingress endpoint. The system is now stable.