It looks like the hosting facility was messing around with their switching setup during that time, hopefully they are done doing this (they have done this a few times in the last couple of weeks around the same time of day). Nothing in the server logs showed problems on our end so was likely just routing issues with them modifying their switching setup.