netcraft uptimes

Top Page
Attachments:
Message as email
+ (text/plain)
Delete this message
Reply to this message
Author: plug@arcticmail.complugarcticmail.com
Date:  
Subject: netcraft uptimes
* On Wed, Nov 15, 2000 at 10:29:14PM -0700, wrote:
> >
> >
> > > Additionally, NT4 uptimes cycle back to zero after 49.7 days, and give
> > > timestamps exactly as if the machine had been rebooted at this precise
> > > point, while HP-UX, Linux, Solaris and recent releases of FreeBSD also cycle
> > > back to zero after 497 days. NT4 SP5 sometimes gives unreliable data,
> > > appearing as a "swarm of bees" effect on a graph
> > * On Wed, Nov 15, 2000 at 08:37:02AM -0700, Trent Shipley wrote:
> > > The fundamental reason all are running *nix seems to be that no other OS
> > > reports uptime . . . even many *nixs with reputations for reliability
> > > evidently don't report uptime. (However note that NT5 *does* report uptime.
> >
> >
> > I *TOLD* you guys that my FreeBSD box had 498 days of uptime,
> > but no one would believe me!
> >
> > Can you EVEN IMAGINE telling someone accustomed to an M$ environment,
> > "We consistently run into a little snag EVERY SINGLE TIME the bloody
> > system has been up for more than 497 days. It's such a pain."
> >
> >
> > D
> >
> >
> Where I work at they need to reboot the NT Server once a week which means
> everything goes down for at least 1 hour. So whenever I'm ask how often I
> have to reboot my web/database/mail/ftp/news server I get very
> disbelieveing looks when I tell them "I don't know, had to be at least
> several months ago". I'm then told that I most likely don't have the same
> amount of traffic which I countered by pointing out how many of the
> employees there use one of my server in one form or another.


Heh. This is great! I think the best (or worst) NT story that
I've heard was about a site (production, cluster of 20+ NT boxen)
that rebooted each NT box every 90 (ninety) minutes on a scheduled
rotation checklist, similar to a Taco Bell toilet check rotation
checklist. Fortunately, the site had a 24x7 operators to perform
scheduled tasks like this. Of course, one could prolly use FreeBSD,
some X-10 relays and some crontab entries to give the NT boxen a
good swift kick in the arse every 90 minutes... :)


D