On Friday, March 17, we experienced a site-wide outage from 3:52–6:30pm PDT. Our engineering team spent the majority of the outage time identifying the source of the problem, which we discovered was a network file storage system. Although we have extensive monitoring systems in place, the issue with the network file storage system was not immediately detected. As a result, our investigation focused on secondary errors before arriving at the root cause. We are making corrective adjustments to our monitoring of this system, so that any future issues can be diagnosed more rapidly. Additionally, we are enhancing our ability to communicate with clients during a major outage such as this one.