Something is going on right now on the VPS where the production instance of http://kdm-manager.com lives. The system is currently inaccessible, but the webserver is still forwarding requests, so it's not down. Could be some kind of utilization thing.
[Update]: as of about a minute ago (Tue Aug 30 15:04:55 EDT 2016) we're back. The system didn't go down, but there was some kind of memory utilization issue that prevented the webserver (nginx) from responding to proxy requests. I'm looking at it presently.
[Update]: still no solid ideas about root cause, but it looks like only two (maybe three) users were potentially affected.