Stuff Happens
From my hosting company:
There appears to be a major routing issue on one of the backbone networks (Global Crossing) that services our location. It is affecting many people trying to reach our network.
Although this is not a problem located on or related to our network, we will monitor it carefully and post more information if we obtain it.
Fortunately, such outages tend to be very short, and very rare.
2 comments
Fortunately, he was right: the outage was short. I tried to post near the apparent beginning of the outage; less than 20 minutes later, I was able to post.
What I want to know is this: how can such a significant portion of the net go down for even that long? I thought DARPA designed the original network to be fail-soft in the face of multiple outages. (Actually, I guess this was fairly soft.) I hope and presume the military networks did not experience this broken backbone.
The original network couldn’t be compromised like this, but since it went commercial and people stopped maintaining their own DNSs, but farmed the job out, and reduced the redundancy, it can happen.
It looked like a routing problem – you got so far and starting looking for the DNS for the end point, and it wasn’t there. It takes a bit for other routers to notice and establish a path to the working DNS.
At least the NFS status site was available to tell us what was going on.