I've read of a Linux box with an uptime of over three years
on the net at some stage but can't for the life of me find that
link again ... but here's a few other interesting bits & pieces :)

http://uptime.netcraft.com/up/hosted...216.239.63.255
496 days for the top-box doesn't look too shabby ;)


[edit]
Googled some more :D
http://counter.li.org/reports/uptimestats.php
[/edit]

Cheers,
Tink

hehe that rocks, having the max uptime as a record (496 days, after that the time gets reset on linux if im correct).
my longest uptime is about 35 days, then I shut off my computer myself

I aim for setting my maintenance periods every 3 months. I don't know what my longest would have been. While I was in school it was always sub-four-months 'cause I had to move... after school... I ran a webserver/firewall in RH7.2 that was up for quite a long time... I would guess 6 months at a time....

Now? My current box has been up over 42days. It runs my ssh server, vncservers for 5-8 users, firewall, ftp, dhcp, and uh various desktop managers for me when I'm at home... *shrug* not a busy box for an AMD1.4 w/256MB (frigging blew-up the other 256. :@ ) The services get restarted as required for upgrades etc. but I usually leave hardware and kernel upgrades for each quarter.

I think your question could be answered "indefinitely". If you had a constant supply of power and there were no apps running on the box... it would chug away until the hardware gave out. BUT... that brings in another question... if you were running on a machine without moving parts... wow... it could run until we die. :P

We just installed a RH 9.0 box at work. Once I had everything setup, we put in a closet with just the power and ethernet plugged in. Was up for almost a month, when I had to take it down to put in a second nic. (Front door for the web site) I don't see any reboots anytime soon, can't upgrade the kernel, 2.4 is needed for the hardware raid and intel gigabit ethernet. :)


Chris Van Brederde
International Logistics Systems
http://www.intlogsys.com

(server: 2.4GHz P4 w/ HT, 1024MB dual PC3200, 2xWD SE 120GB (mirrored), 2xWD SE 40GB (mirrored), WD 20GB (for the OS),.)

62 days for me, could have been longer but I was going out of town for 2 weeks, and figured there was no point in paying for a bunch of kilowatts just to keep the uptime streak alive. Second best was 59 days, cut short due to a brief power blackout.

The main thing though is that my Windows machine would regularly need to be rebooted just about every 1 or 2 days. Moving to Linux was easily one of the best decisions I've made -- J.W.

I just had a week or so till today. Someone decided to pass and didn't make it. One of the two hit the utility pole. No power, no UPS monitor program either. I bet they are hurtin though. Went through a deep ditch, !threw! the pole, then threw a guard rail, and then threw a fence. I don't think they hit any cows though. That pole looked like toothpicks.

He he he he, the telephone lines look like spaghetti. Wouldn't want to fix that mess of wires.

I need to get something for my UPS but nothing available for Gentoo, yet.

Later



Gentoo's not in the dictionary?

10:30am up 404 days, 18:31, 1 user, load average: 0.00, 0.00, 0.00

- but the machine is a laptop, so poweroutages are coped with by the internal battery... I'm reluctant to upgrade the distro and loose the uptime..

Arg. This thread was bad luck. About 5 days after I read it, I realized I had to make a BIOS adjustment (stupid stupid stupid mistake, my 1.4GHz machine was running at only 1GHz see my other thread)...

Anyway, reset my counter and start from scratch. :/

Taking this slightly off-topic, but not far:

With the new 2.6 kernels you get ACPI support, so you can so a software suspend (sleep mode). If I am right, this puts the contents of your RAM into a reserved area of your SWAP and sets a flag at the beginning that says "hey, if you see this flag, then copy the contents of SWAP back to RAM rather than normal booting". If this is the case, what does this do to uptimes? Does the uptime continue where it left off regardless of the fact that the machine was stopped, or is it cleared as in a normal reboot?

Oh, and my biggest uptime was about 3 weeks. I left one of my machines running at home (by accident) when I went back to Uni. Didn't return home for a while. Currently I don't leave my machine running all-the-time as it is damn noisy and it lives in my bedroom. Sleep and computer-running are, at the moment, mutually exclusive.

wow, are u not worried that it will drastically shorten the lifespan of your laptop?

my max linux up time was 42 days, came home from work 1 day, the screen was dark, moved the mouse, windowmaker turned back on, clock was frozen at 14:22, mouse wouldnt move, keyboard wouldnt respond, hard booted, ext2 hdd, corrupt home account
installed freebsd, max uptime 219 days before i moved out

My desktop computer (Slackware, although it would have been running 9.0 at the time) stayed on with me for like 3 months, I would take a guess and say 2 and a half months.

[root@cache root]# uptime
07:37:05 up 149 days, 19:12, 2 users, load average: 0.11, 0.09, 0.09
[root@cache root]#

This is on a server that handles around 1.5 million hits a day on the web and also has about a terabyte of transfer a day, I have been through 2 DoS attacks since the last reboot, the last one I was handling over 8000 concurrent connections at a time. I don't have any windows machines that can boast that :-)

Don't feel bad, our server's got reset...I was closing the closet door it hides behind, and the door push on a cord that push on the switch on the power strip that it is pugged into. Still plugging the boss man of a real UPS for it. But at least I learned to move the strip so that wouldn't happen again.

Chris V.

http://www.intlogsys.com

i'm curious, whats the site that its getting hammered so much? it it part of a cluster? seems a little 'light' for 1 machine to be at it!
how the performance after that amount of up time? thats my biggest question that i have been researching recently ( freeBSD seems to be kicking Linux ass with that 1)