Among the multitude of problems that have plagued this process, we have discovered yet another one.
This time, it’s not actually our fault. It seems that the nifty new Juniper we deployed there is suffering from a hardware failure in the form of a memory failure which is causing a number of odd behaviors (such as forwarding packets in ways that don’t agree with the forwarding table).
We are working to resolve this latest issue as quickly as possible and will continue to work to bring this site back up to full capacity as soon as possible. We apologize (again) for any inconvenience.