CPU workunit availability
log in

Advanced search

Message boards : Windows : CPU workunit availability

Author Message
altheogre
Avatar
Send message
Joined: 28 Feb 13
Posts: 5
Credit: 12,270,755
RAC: 0
Message 16656 - Posted: 30 May 2013, 9:06:36 UTC

Hi,

For the last couple of days I have not been able to get CPU work. Is there currently none available?

30/05/2013 09:50:46 | Collatz Conjecture | update requested by user
30/05/2013 09:50:49 | Collatz Conjecture | Sending scheduler request: Requested by user.
30/05/2013 09:50:49 | Collatz Conjecture | Requesting new tasks for CPU
30/05/2013 09:50:51 | Collatz Conjecture | Scheduler request completed: got 0 new tasks
30/05/2013 09:50:51 | Collatz Conjecture | Message from server: NVIDIA GPU: Upgrade to the latest driver to process tasks using your computer's GPU
30/05/2013 09:50:51 | Collatz Conjecture | No tasks sent

Cutchet Salvador
Send message
Joined: 22 Jan 13
Posts: 9
Credit: 493,055
RAC: 0
Message 16659 - Posted: 30 May 2013, 11:48:39 UTC - in response to Message 16656.

I am in the same situation, today it is the second day without work.
Alone work for GPU not for CPU.
Greetings

Profile Pooh Bear 27
Avatar
Send message
Joined: 1 Aug 10
Posts: 54
Credit: 108,227,920
RAC: 0
Message 16660 - Posted: 30 May 2013, 12:28:07 UTC

If you look in the number crunching I posted about it having issues with check pointing correctly. If you look at the Applications page the application has been pulled.

The admins replied they knew there was an issue with the code and probably pulled it until they could create a fix.

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 2
Message 16662 - Posted: 30 May 2013, 15:03:00 UTC

Correct. I pulled the 4.xx app for CPUs due to the checkpoint issue and put back the previous 2.xx version. After reading your posts I did some investigating and sure enough, some BOINC developer had removed SSE checking. I just put it back in so it should be working again now. My guess is that the apps that used to have multiple versions for different CPU types (SETI, EINSTIEN, MW, etc.) have since merged the code into a single app and DA therefore removed the SSE check without asking anyone else whether they were using it. Thankfully, there is still a SSE3 check so it was easy to add SSE since the code is exactly the same (minus the "3").

Profile kmanley57
Send message
Joined: 1 Apr 12
Posts: 6
Credit: 32,623,691
RAC: 0
Message 16667 - Posted: 30 May 2013, 23:22:37 UTC

Thank you! You made my day with getting new work.

Cutchet Salvador
Send message
Joined: 22 Jan 13
Posts: 9
Credit: 493,055
RAC: 0
Message 16671 - Posted: 31 May 2013, 7:56:46 UTC - in response to Message 16662.

Thank you, already I have work.
His rapidity of response is incredible!
Greetings

Cutchet Salvador
Send message
Joined: 22 Jan 13
Posts: 9
Credit: 493,055
RAC: 0
Message 16711 - Posted: 5 Jun 2013, 8:18:15 UTC - in response to Message 16662.

After of the stop of the server, I have not received WU'S again for the CPU.
Is it normal?
Greetings

Profile mikey
Avatar
Send message
Joined: 11 Aug 09
Posts: 3242
Credit: 1,685,568,533
RAC: 6,385,245
Message 16715 - Posted: 5 Jun 2013, 11:06:24 UTC - in response to Message 16711.

After of the stop of the server, I have not received WU'S again for the CPU.
Is it normal?
Greetings


Does your other project have lots of work, Boinc will only get new work as long as it thinks it can finish it before the deadline, since you have two projects and Collatz was unavailable you may have loaded up on work from the other project and will get more Collatz work as your current cache runs down a bit.

Cutchet Salvador
Send message
Joined: 22 Jan 13
Posts: 9
Credit: 493,055
RAC: 0
Message 16716 - Posted: 5 Jun 2013, 11:51:17 UTC - in response to Message 16715.

Thanks Mikey, I will abort the works and I will only collatz.
Thank you and greetings

Profile mikey
Avatar
Send message
Joined: 11 Aug 09
Posts: 3242
Credit: 1,685,568,533
RAC: 6,385,245
Message 16729 - Posted: 6 Jun 2013, 11:16:52 UTC - in response to Message 16716.

Thanks Mikey, I will abort the works and I will only collatz.
Thank you and greetings


You don't HAVE to do that, Boinc should do it on its own now that Collatz is back up again. Another possible solution is to make your cache size smaller so you don't get quite so much work when you do get work. Another option is to use the percentages to make Boinc work more 'your way' instead of just the 'default way'. If you go to a projects web page and then under your account change the 100% default to 0%, then it will only get work when a project with a higher percentage is out of work. I have several 'backup projects' I do that with. I have my projects I am working on now but sometimes they run out of work, or are down. When that happens my backup projects kick in but since the percentage is set at 0% then only get enough units to keep the cpu's or gpu's busy but don't get any extras. Then when they need more work they check the projects with a higher percentage first, then get work for the project with the 0% if the others still don't have any work.


Post to thread

Message boards : Windows : CPU workunit availability


Main page · Your account · Message boards


Copyright © 2018 Jon Sonntag; All rights reserved.