Some of you may have noticed that our service was down for about 15 minutes this afternoon. We were here working on it, so we noticed it immediately and traced the problem to one of our three online service providers. They had been alerted internally and were already working on the problem, so service was restored very quickly in this case. For those who want to know more, read on...
Our media authorization network (MAN) provider (log in, customer accounts, credit card transactions and security) suffered a technical failure to one of their routers (an Internet address and traffic directing machine) this afternoon that affected all their customers including us — a first. The machine was saturating our domain and others with thousands of requests per minute, similar to a "denial of service" attack.
They recovered and repaired the situation immediately by rebooting the confused machine. After that hos.com was again accessible but slow. At the time of this post, everything is back to 100% speed and efficiency. We put up a "down for maintenance" notice once we became aware of the outage, then went back to normal when the issue was resolved.
The Internet may not be "a series of pipes" (Sen. Ted Stevens of Alaska) but it is run by machines, software, and people. And sometimes they don't behave as instructed and need a timeout.