Connection latency between 11:51am and 12:08pm EDT, and again briefly at 2:23pm EDT, caused some slow page loads and roughly 5 minutes of downtime today. Both incidents were caused a problems with our upstream hosting provider, AWS. The first problem resulted in connection failures to session data stored in cache, and the second was a very brief failure to connect to our web servers.
We are capable of continuing service when a single data center is down, but in this case, the problems seem to have impacted the entire region. Service has returned to normal and all systems remain stable on our side of things.