Re: [SLUG] Record uptime for Linux

From: robin (robin@roblimo.com)
Date: Sun May 05 2002 - 15:20:25 EDT


R P Herrold wrote:

>On Sat, 4 May 2002, Bill wrote:
>
>
>
>>t seems to me that I have read of Linux boxes with uptimes of years. Does
>>anyone have a link to a web page about this topic?
>>
>>
>
> http://bopper.wcbe.org/uptime.txt
>
>prior run was 900+ days, current run is at 729 days
>
>"The administrator has not physically touched
>it or its keyboard since 1997. It has no monitor."
>

And a busted hard drive forced them to reboot.

Now I'm going to scare you: I was at my friend Joe's place (he's the
Linux clonebuilder in Baltimore I mentioned a few days ago) one night,
scarfing frozen Vodka shots, and he yanked the hard drive ribbon cable
off of a motherboard that was running FreeBSD. It kept running. He
plugged the hard drive back in, and it kept running.

In theory, you could make a motherboard running FreeBSD (or most other
BSDs) go on until it had an onboard failure if you had redundant or
hot-swappable power supplies and hot-swapped hard drives well before
they hit their manufacturers' MTBF specs. And a motherboard with a
conservative CPU/RAM and adequate cooling could have a potential
lifespan of 100 years or more.

This is an example of *BSD's technical superiority/reliability over Linux.

Okay, it's a carny trick that isn't terribly relevant in real life work
situations, but if your goal in life is a computer that'll run long
enough to control a starship running at sublight speeds, you're probably
better off running BSD Unix than Linux.

- Robin



This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 17:42:41 EDT