Every time I see a story about someone having to reboot their machine at least every 24 hours just makes me chuckle.

My XP Pro machine at home routinely is on 2 week uptimes. Normally the only time I have to reboot is when I feel like it based on a new driver and I want to clear out the memory or a hardware add (which is way too frequent). This machine has so much crap software loaded on it and then uninstalled that the registry must be a royal mess, yet it keeps on trucking. I haven’t had a blue-screen or spontaneous reboot since the last formatting I did (which was done due to hard drive movements because of a failed drive). Athlon 1800+, Abit board, 768MB RAM, 6 Hard Drives (at last count 🙂 ), DVD+-R, DVD-ROM.

My machine at work (an XP Pro laptop that has to come home with me, hence can’t stay on 24 hours at work but is on at home from about 10am Mon-4pm Fri) has yet to crash on me during a workday/night (12 hour graveyard shifts). This is using Java Based network monitoring apps (NetCool), “wonderful” Lucent Naviscore monitoring (anybody that uses that KNOWS how much of a hog it is), times of 10-15 telnet sessions, Citrix sessions, email and who knows what else at different times. I have not had to reboot during a shift since given this machine 3 months ago, yet I hear my co-workers complaing quite a bit about having to reboot their machine in the middle of a maintenace (notice that the hearing part isn’t second-hand…it’s literally hearing them yell across the room that their machine crashed again. Their machines are identical hardware running Linux (Mandrake I believe) that is running the same apps as the XP installs (or as close as can be found). IBM Thinkpad, P4M 2.0GHz, 768MB RAM.

When I did contracted onsite help desk support for a 500+ user company using Win98SE, I did a survey for 2 weeks just to see what real-world shows vs perceptions. I put into the network login script, a counter, which basically added up the number of times each machine (not user, but machine MAC address) was logged onto the network. This gave me a relatively good sample to figure out how many times each machine was rebooted during the 2 weeks. At the beginning I had gotten permission to kick every MAC address off of the server so that I had a complete number in regards to machine logins. Results were that only a small handful (less than 15) of the logins were the result of crashes. The majority were due to users who turned off machines at the end of the day and then back on again when they got into work. Frankly, I was surprised with the number…I did expect it to be a little higher. This sample was done with typical office people..very few of them were really computer adept, just general users that did their entire jobs on the machine. Maybe my company just built really good machines for our clients… 🙂

Nate