From about 5am to 9am PT, we experienced an increased number of failed requests to our API, which caused the site to be for some users, and for some admin tools to stop working. The issue was related to a failure of one of our systems to scale to additional resources, causing the system to be overloaded. We have identified and fixed the issue.