Advanced search

Message boards : Server and website : Please cancel workunits if you have too many

Author Message
Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 20972 - Posted: 15 Apr 2011 | 8:11:25 UTC

This will allow other people to get them.

gdf

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 20973 - Posted: 15 Apr 2011 | 8:26:05 UTC - in response to Message 20972.
Last modified: 15 Apr 2011 | 8:28:49 UTC

What about doing something about these hosts? Imagine the bandwidth you'd save.

http://www.gpugrid.net/results.php?hostid=67477
http://www.gpugrid.net/results.php?hostid=98004
http://www.gpugrid.net/results.php?hostid=89811
http://www.gpugrid.net/results.php?hostid=58604
http://www.gpugrid.net/results.php?hostid=48006
http://www.gpugrid.net/results.php?hostid=97398
http://www.gpugrid.net/results.php?hostid=60046

These are but a few.
____________
Radio Caroline, the world's most famous offshore pirate radio station.
Great music since April 1964. Support Radio Caroline Team -
Radio Caroline

Toni
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 9 Dec 08
Posts: 1006
Credit: 5,068,599
RAC: 0
Level
Ser
Scientific publications
watwatwatwat
Message 20975 - Posted: 15 Apr 2011 | 11:50:39 UTC - in response to Message 20973.

We trust the back-off mechanism of boinc...

Jacob Klein
Send message
Joined: 11 Oct 08
Posts: 1127
Credit: 1,901,927,545
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 20997 - Posted: 18 Apr 2011 | 13:55:48 UTC - in response to Message 20975.

Is it possible that something could be done to investigate the errors and change the applications so that they fail less?

That would be the ideal solution.

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 20998 - Posted: 18 Apr 2011 | 14:29:06 UTC - in response to Message 20997.

Nah, they can't be bothered. Same with news and updates about the project and science. Just keep crunching and hopefully donating plus pay you electric bill and keep up with the latest cards and we'll give you a whole bunch of meaningless credits. Having said that most BOINC projects are the same, a lot of take and very little give.


____________
Radio Caroline, the world's most famous offshore pirate radio station.
Great music since April 1964. Support Radio Caroline Team -
Radio Caroline

Otis11
Send message
Joined: 2 Aug 09
Posts: 21
Credit: 197,088,189
RAC: 0
Level
Ile
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 21014 - Posted: 20 Apr 2011 | 2:40:24 UTC - in response to Message 20997.

IDK if this is the same situation, but I know even on a contained supercomputer where nothing is overclocked, they have occasional errors... sometimes it's due to errors in the code, and sometimes it's due to unexpected situations in the science they are investigating. The first they try to prevent as much as possible because it hurts they're performance (just like it hurts our points) and they need more performance than we can give them, and the later just comes with the territory.

As a programmer I actually have to hand it to them. Programming applications like this is difficult enough. Pushing it across a distributive system just introduces more potential for difficulties.

ftpd
Send message
Joined: 6 Jun 08
Posts: 152
Credit: 328,250,382
RAC: 0
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 21016 - Posted: 20 Apr 2011 | 9:15:30 UTC

20-4-2011 11:13:23 GPUGRID work fetch resumed by user
20-4-2011 11:13:25 GPUGRID Sending scheduler request: To fetch work.
20-4-2011 11:13:25 GPUGRID Requesting new tasks for CPU and GPU
20-4-2011 11:13:27 GPUGRID Scheduler request completed: got 0 new tasks
20-4-2011 11:13:27 GPUGRID Message from server: No work sent
20-4-2011 11:13:27 GPUGRID Message from server: No work is available for ACEMD beta version
20-4-2011 11:13:27 GPUGRID Message from server: No work is available for Long runs (8-12 hours on fastest card)
20-4-2011 11:13:27 GPUGRID Message from server: ACEMD beta version is not available for your type of computer.


Server has enough wu's availabel?

Already known problem?

All my computers cannot download!
____________
Ton (ftpd) Netherlands

Toni
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 9 Dec 08
Posts: 1006
Credit: 5,068,599
RAC: 0
Level
Ser
Scientific publications
watwatwatwat
Message 21017 - Posted: 20 Apr 2011 | 10:03:54 UTC - in response to Message 21016.

It's part of a known problem, but we now need to check if the latest changes fixed the scheduler. We have ~100 WUs on the long and ~100 on the short queue, so please attempt an update. Do you still get no work?

ftpd
Send message
Joined: 6 Jun 08
Posts: 152
Credit: 328,250,382
RAC: 0
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 21018 - Posted: 20 Apr 2011 | 10:13:37 UTC - in response to Message 21017.

Toni,

Finally i got one wu for my gtx480-machine. I will try the rest also now!
The normal download is two wu's for this machine.

Second change is NOT OK!


____________
Ton (ftpd) Netherlands

Post to thread

Message boards : Server and website : Please cancel workunits if you have too many

//