This is becoming almost traditional: just before I embark on a trip, my otherwise ultra-reliable server crashes while performing some routine operation.
Back when dinosaurs still roamed the Earth and the Internet was still an exotic novelty for most people, my wife and I spent an unforgettable few weeks driving around the United States.
My server has been in existence for more than two years by then. No Web site yet… that came later that year. (Actually, I already had a rudimentary version up and running courtesy of my Internet service provider, but I digress.) The server was “connected” to the Internet via dial-up; it dialed regularly (once an hour; my wife and I still remember the little touch-tone “song” that we heard many times a day) to exchange e-mail, and it dialed automatically if I tried to visit a Web site, for instance. Otherwise, the modem line was open for dial-in… and I planned to dial into the server regularly while traveling.
Except that on the morning of our planned departure, the server crashed. Not a minor crash either… its hard drive died. Fortunately, not an instant death, and I was able to obtain a replacement drive and rescue all data.
Though it was still a genuine, no-kidding, bona fide kernel panic, today’s crash was a little less dramatic. I was merely testing streaming video from the server, as I always find it useful to be able to watch an Ottawa newscast in the evening when I am out of town. Instead of streaming the darn video, the video capture driver (which I’ve been using, completely incident free, for many years) crashed, taking the server down. Most unsettling!