Rebuilding the cache appears to have resolved the CDN problem, but we'll continue to monitor the situation until we're sure things are fixed.
Posted over 2 years ago. Mar 17, 2015 - 01:53 UTC
We've identified a rogue web node, which corrupted our CDN's cache. That node has been removed from the cluster, and we're forcing a refresh of the CDN. More updates soon!
Posted over 2 years ago. Mar 17, 2015 - 01:44 UTC
We're currently investigating some partially broken JS, which may affect some users' ability to perform certain actions in the Web UI. Git and Mercurial operations should not be affected in any way. This should be resolved shortly after 01:30 UTC.