Join PrimeGrid
Returning Participants
Community
Leader Boards
Results
Other
drummers-lowrise
|
Message boards :
Problems and Help :
Too many work units keep coming in
| Author |
Message |
|---|
 ClintSend message
Joined: 2 Jan 17
Posts: 2
ID: 479932
Credit: 359,721,893
RAC: 1,170,697
            
|
|
My settings in this project and BOINC manager are to have .5 days work, with up to .5 day extra and I crunch PPS Sieve. I had to abort 36 jobs today due to time deadline. Then it just fetched 72 new work units. My slow little GPUs may come close to completing this work but it looks like I'll have to abort some more in a few days. How do I break this cycle and get reasonable amount of work at a time? Thanks | |
|
Michael Goetz Volunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14011
ID: 53948
Credit: 433,660,774
RAC: 817,079
                              
|
My settings in this project and BOINC manager are to have .5 days work, with up to .5 day extra and I crunch PPS Sieve. I had to abort 36 jobs today due to time deadline. Then it just fetched 72 new work units. My slow little GPUs may come close to completing this work but it looks like I'll have to abort some more in a few days. How do I break this cycle and get reasonable amount of work at a time? Thanks
First of all, you don't have to abort the tasks. The way BOINC works, once you pass the deadline, it automatically aborts any tasks that haven't started yet. If you simply ignore the problem, nothing horrible will happen. (Aborted tasks don't hurt you, PrimeGrid, or anyone else.)
That, of course, doesn't directly answer your question. The general answer to the question "Why does it download so many tasks" is that BOINC gets confused about how long each task takes. This happens when you're running different types of PrimeGrid tasks. BOINC has trouble keeping track of how quickly PrimeGrid X task should run after having run PrimeGrid Y task at a different speed.
I typically run with 0 days (because I'd rather be the discoverer of a prime than a double checker), and that will usually cause BOINC to download just a single task at a time. Usually.
"0" actually means "3 minutes". I intended to download a single GFN 15 for testing the other day. They're very short. I was set for 0 days. 12 tasks got downloaded. They're short, but they're not THAT short. They take about a minute each. Normally, however, a setting of 0 will get you just a single task. If you have a 24/7 and reliable connection to the Internet, this is what I recommend. BOINC will download a new task 2 to 3 minutes before the last one ends, so you don't need to worry about the computer running out of work.
____________
My lucky number is 75898524288+1 | |
|
ClintSend message
Joined: 2 Jan 17
Posts: 2
ID: 479932
Credit: 359,721,893
RAC: 1,170,697
|
|
Ok, thanks for the feedback. Yeah I process cpu and gpu tasks but it was typical to get maybe a dozen tasks combined. When I started seeing 70+ it looked like a problem. Good to know it doesn't affect the project when having to reassign WU's at the expire date. Typical GPU task is expected 4hrs on my system so getting swamped was alarming. Hopefully I'm not spending hours crunching expired ones. | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14011
ID: 53948
Credit: 433,660,774
RAC: 817,079
|
Hopefully I'm not spending hours crunching expired ones.
That won't happen, since if it's expired, it will automatically have been cancelled before it starts.
____________
My lucky number is 75898524288+1 | |
|
Message boards :
Problems and Help :
Too many work units keep coming in |
