Uptime

Discussion of chess software programming and technical issues.

Moderators: hgm, Rebel, chrisw

User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Uptime

Post by sje »

Due to power interruptions, it's uncommon that any of my desktop machines is up for much more than a month or two at a time.

But for my old battery powered HP notebook running Linux and working on the Perft(14) project:

Code: Select all

sje@nancy:~/Arena/Symbolic$ uptime
 00:29:03 up 152 days, 23:52,  3 users,  load average: 2.00, 2.01, 2.05
Joost Buijs
Posts: 1563
Joined: Thu Jul 16, 2009 10:47 am
Location: Almere, The Netherlands

Re: Uptime

Post by Joost Buijs »

Here in the Netherlands the situation is somewhat better, in my city we've had maybe 2 or 3 power outages over the last 15 years.
I've had Windows machines (automatic updates disabled) running for more then a year without any interruption, so who's telling Windows is not stable?
matthewlai
Posts: 793
Joined: Sun Aug 03, 2014 4:48 am
Location: London, UK

Re: Uptime

Post by matthewlai »

sje wrote:Due to power interruptions, it's uncommon that any of my desktop machines is up for much more than a month or two at a time.

But for my old battery powered HP notebook running Linux and working on the Perft(14) project:

Code: Select all

sje@nancy:~/Arena/Symbolic$ uptime
 00:29:03 up 152 days, 23:52,  3 users,  load average: 2.00, 2.01, 2.05
09:27:40 up 361 days, 16:16, 1 user, load average: 0.12, 0.04, 0.05

That's my AWS server, so I guess that's cheating a bit.
Disclosure: I work for DeepMind on the AlphaZero project, but everything I say here is personal opinion and does not reflect the views of DeepMind / Alphabet.
mvk
Posts: 589
Joined: Tue Jun 04, 2013 10:15 pm

Re: Uptime

Post by mvk »

Joost Buijs wrote:Here in the Netherlands the situation is somewhat better, in my city we've had maybe 2 or 3 power outages over the last 15 years.
I've had Windows machines (automatic updates disabled) running for more then a year without any interruption, so who's telling Windows is not stable?
Quite rare here indeed. I never understood all those ads for surge protectors and battery backups in US magazines. Just a random system I normally use for opening prep:

Code: Select all

11:10AM  up 872 days, 12:06, 1 user, load averages: 0.06, 0.08, 0.07
The last reboot was to add an SSD.
[Account deleted]
User avatar
hgm
Posts: 27794
Joined: Fri Mar 10, 2006 10:06 am
Location: Amsterdam
Full name: H G Muller

Re: Uptime

Post by hgm »

It makes a big difference when the power lines are under ground. Not much worry about lightning. (The truly high-voltage long-distance lines are usually well protected against lightning, by having grounded wires just above them.) Most power failures are due to stupid workmen digging in the road with heavy equipment, without properly studying the infra-structure maps first.

I can imagine that on rocky terrain having subsurface cabling is prohibitively expensive. It is one of the few advantages of building in a swamp. :lol:
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Above and below

Post by sje »

Nearly every power outage in my vicinity is caused by storms which topple trees which in turn topple power lines. Wind, snow, and ice conspire in the winter months for power outages that sometimes last for days.

Routing power lines below the ground is done mostly only for some government facilities at the federal level because of the great expense of running cabling through rocks.

Perhaps one out of a dozen outages is caused by a motor vehicle crashing into a utility pole.

The last power outage here was 56 days ago. The longest outage in the past decade was about four days with no telephone service for eight days.

Supplying battery power for the desktops is outside my current budget. However, in the Old Days (first half 1990s) when I had but one machine to build some five man tablebases, the months-long runs were supported and saved a couple of times with uninterruptible power supplies.
User avatar
mhull
Posts: 13447
Joined: Wed Mar 08, 2006 9:02 pm
Location: Dallas, Texas
Full name: Matthew Hull

Re: Uptime

Post by mhull »

sje wrote:Due to power interruptions, it's uncommon that any of my desktop machines is up for much more than a month or two at a time.

But for my old battery powered HP notebook running Linux and working on the Perft(14) project:

Code: Select all

sje@nancy:~/Arena/Symbolic$ uptime
 00:29:03 up 152 days, 23:52,  3 users,  load average: 2.00, 2.01, 2.05
iBook 900 MHz PowerPC G3, 640 MB SDRAM

12:55 up 1424 days, 19:18, 2 users, load averages: 0.40 0.12 0.17

Top that. :)
Matthew Hull
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Re: Uptime

Post by sje »

mhull wrote:iBook 900 MHz PowerPC G3, 640 MB SDRAM

12:55 up 1424 days, 19:18, 2 users, load averages: 0.40 0.12 0.17
That's much better than my all-time record of some 500 days. I have read that a Linux system should be rebooted at least once a month to once a year to ensure that all updates are thoroughly integrated into the running image, so I haven't had an uptime longer than a year for a while.

If that iBook is running OS/X 10.4, then it's been more than four years since it has needed a reboot due to a system update. I have a 2006 PowerBook running OS/X 10.7 and it might have had a 2,000+ day uptime if it were running on its battery. Too bad that its battery blew up a few years ago.

Code: Select all

melissa:Symbolic sje$ uptime
13:24  up 56 days, 10:27, 3 users, load averages: 2.69 2.68 2.62
Perft(14) work unit 325 is 77% done on Melissa while work unit 366 is 46% done on Nancy. The most recently started work unit, number 392, is being processed by Rebecca and is 18% complete:

Code: Select all

sje@rebecca:~/Arena/Symbolic$ uptime
 13:43:00 up 55 days, 14:46,  3 users,  load average: 8.05, 8.04, 8.04
Others may have longer uptimes, but all my machines are running full loads.
mvk
Posts: 589
Joined: Tue Jun 04, 2013 10:15 pm

Re: Uptime

Post by mvk »

hgm wrote:I can imagine that on rocky terrain having subsurface cabling is prohibitively expensive. It is one of the few advantages of building in a swamp. :lol:
One left-pondian once explained the infra quality difference as more of a capitalism vs socialism phenomenon. A litmus test would be to compare Florida with Sweden.
[Account deleted]
User avatar
mhull
Posts: 13447
Joined: Wed Mar 08, 2006 9:02 pm
Location: Dallas, Texas
Full name: Matthew Hull

Re: Above and below

Post by mhull »

sje wrote:Nearly every power outage in my vicinity is caused by storms which topple trees which in turn topple power lines. Wind, snow, and ice conspire in the winter months for power outages that sometimes last for days.

Routing power lines below the ground is done mostly only for some government facilities at the federal level because of the great expense of running cabling through rocks.

Perhaps one out of a dozen outages is caused by a motor vehicle crashing into a utility pole.

The last power outage here was 56 days ago. The longest outage in the past decade was about four days with no telephone service for eight days.

Supplying battery power for the desktops is outside my current budget. However, in the Old Days (first half 1990s) when I had but one machine to build some five man tablebases, the months-long runs were supported and saved a couple of times with uninterruptible power supplies.
For critical work to continue for short outages, a UPS might do at least until you get your alternate power up and running. I have a Honda 3000ui generator (2800 watts continuous capacity) which ran for 10 days without interruption until the mains were restored. It's very quiet and provides clean power for sensitive electronics. Not cheap though.
Matthew Hull