Forum Moderators: bakedjake
Most of those boxes are BSD/OS systems, so you have to wonder what makes BSD/OS (by BSDi) so popular in the long-uptime arena.
Generally, folks who buy BSD/OS are of the mentality that's perfect for big uptimes. Get the hardware, install the OS and apps, forget that the box existed unless their syslog partition fills up. If they were tinkerers, they wouldn't be buying BSD/OS, they'd get a free OS so that they could .. er.. well.. tinker.
Linux/FreeBSD/NetBSD/OpenBSD folks, for their own reasons, are often rebooting their systems.
For example, those who pick up OpenBSD are security conscious, so they're keeping an eye on the security warnings that come out. When one does, apply patch (and/or recompile) and reboot.
I often like to think of linux users as the ultimite in tinkerers. Most of my interaction with linux users involve walking into their cube or house and they have at least one linux PC all busted apart and *RUNNING* while exposed. Before you get upset at this generalization, I do know linux folks who are at the other end of the spectrum. They're just closet NetBSD users. :)
Once-upon-a-time, my vanity domain (running NetBSD) was running with an uptime of over a year, and over a 3-year period encountered three reboots -- two of which were >12-hr power failures (outlasting the UPS). Now I'm paying more attention to security issues, so I fix'n'reboot everytime something comes out that affects ssh (which seems to be weekly these days).
So while these stats sure make me feel warm'n'fuzzy in the linux-vs.-BSD banter, I have to admit that this information is somewhat biased to a particular mindset and doesn't really reflect system stability.
Rob++
I've certainly had SunOS boxes last well over that, and I can probably dredge up some memories of Solaris servers that I've nurtured taht long.
Rob++
Rob++
