[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Uptime: The Movie
I realize that there may yet be more who want to get their inputs
in on the uptime question, but we can't wait forever, you realize.
But first, some moralizing:
When I first posted the request, I was a little hesitant, because
I thought you would all think I was weird. I mean, I have avoided
rebooting a machine that had been up a long time, just because I
didn't want to break its record. However, from the responses that
came back, I can see that I am not alone. Sorry about the bandwidth,
but we can have a little fun along the way, can't we?
Anyway, on with the summary:
>From s9274@srl1.LANL.GOV Mon Oct 21 12:14:49 1991
Our server, OSSABAW (a 3670), is currently at 6 weeks, 6 days, 7 hours.
I think we've beat that in the past, but I can't verify it.
John Krieger (s9274@srl1.lanl.gov)
Westinghouse Savannah River Company
----------------------------
>From Dave@PHOENIX.SCH.Symbolics.COM Mon Oct 21 12:50:45 1991
METHUSELAH 30 weeks 2 hours 40 minutes 50 seconds
----------------------------
>From pao@ai.mit.edu Mon Oct 21 13:26:22 1991
Well, that just beats my record of 28 weeks 6 days 20 hours 40 minutes
6 seconds from 19 February to 10 September of this year. I was shut
down because of an A/C failure in our machine room.
- Pat
----------------------------
>From Hornig@RIVERSIDE.SCRC.Symbolics.COM Mon Oct 21 14:39:26 1991
The best I've seen in normal operation was somewhat more than 13 weeks
for a file server here at SCRC. We only shut it down because we were
moving our whole site to a different town.
----------------------------
From: buff@pravda.cc.gatech.edu (Richard Billington)
Months ... many of them. 3 - 5 at least.
(Note - Methuselah still beats this, apparently JDR)
----------------------------
From: Healy@space50.nrl.navy.mil (Liam M. Healy)
As a different measure of age of a boot, I look at words consed
(after all, keeping a machine going is no trouble if you're not using
it). The highest I've seen here is about 500,000,000 words
(from Show GC Status).
----------------------------
From: p2@porter.asl.dialnet.symbolics.com (Peter Paine)
We have had a MacIvory II(fx) stay up for more than 3 weeks so far.
Power cuts are the usual problem.
58% of a year beats our L-mach records flat.
How do other systems standup? ... Sun, Vax, HP, RS6000, NeXT, ...
(Our MacIvory has been up a few weeks at a time, but it is no longer
on the local net. I just checked in at Argonne National Labs, and
their record right now is 154 days [22 weeks] on a lightly used Sun
of some variety. On the Alliant fx2800 I usually use the longest has
been around 45 days, but it was a power outage that did us in [it
is a very new machine] - JDR)
----------------------------
So, as near as I can tell, Methuselah is the winner at 30 weeks,
although I think it is pretty obvious now that Dave has an obligation
to let us know when Methuselah finally ends his streak, so those of
us with machines currently on a streak can know what we need to
shoot for.
Thanks for your indulgence, and I will try to make my next request
a little more serious :-)
John Reading
readingj@cerf.net