Message boards :
Number crunching :
Not getting enough Workunits
Message board moderation
Author | Message |
---|---|
![]() Send message Joined: 24 Feb 10 Posts: 3 Credit: 31,688,828 RAC: 0 ![]() ![]() |
Hello, right now I have 4 work units in my Boincmanager. 3 for my CPU (Xeon E3-1230 V2) and one for my GPU (AMD 7950). The settings in my manager says that he should gather work for at least 5 days. It's ok for the CPU WUs since they take a while to complete, but my GPU runs dry in under a minute and then takes seconds up to several minutes to get the next WU. Boinc is allowed to use 10 GB on my disk and uses ~3GB. Is there a way to get more WUs ? Greetings deKay 04.11.2014 13:11:27 | Milkyway@Home | Computation for task de_83_DR8_rev_8_4_00002_1413455402_7462776_0 finished 04.11.2014 13:51:23 | Milkyway@Home | Requesting new tasks for ATI |
Send message Joined: 2 Feb 09 Posts: 3 Credit: 1,443,819 RAC: 0 ![]() ![]() |
What is your "Resource share" under "MilkyWay@Home preferences"? It feels like that you have zero (marked by ---) Resource share. |
![]() Send message Joined: 24 Feb 10 Posts: 3 Credit: 31,688,828 RAC: 0 ![]() ![]() |
Interesting, this seems to be the first project I encounter, that actually uses/respects the "resource share" option. Thanks, now I get more work units. |
![]() Send message Joined: 4 Jul 14 Posts: 8 Credit: 64,068,552 RAC: 0 ![]() ![]() |
Are there any other possibilities? I have 8 machines right now, some are pulling just enough WUs to skate by, others aren't pulling WUs at all. My resource share is set at 250, and I have BOINC configured to grab 2 days worth of WUs. I'm at a loss at the moment..... |
![]() Send message Joined: 18 Jul 09 Posts: 300 Credit: 303,565,482 RAC: 78 ![]() ![]() ![]() |
1. Unhide your computers. 2. What are your other resource shares set to? |
![]() Send message Joined: 4 Jul 14 Posts: 8 Credit: 64,068,552 RAC: 0 ![]() ![]() |
Computers showing other Resource shares set at 100 (on most of the machines, MilkyWay is the only project running) After a bit of fighting, all of the machines have WUs, but I'm still baffled as to why it's so hard to get them to come down. Computers 584093, 594634, 594866, 592580 all have issues pulling down WUs. Once they complete the ones they have, they'll just show "Ready to Report" until I force an update several times while standing on one leg, wearing a tin foil hat. That, and I'm confused as to why, on the 8-core machines, they only pull down 24 WUs, and my log says they have hit their max. Regardless of what I set the max work buffer, they only get 24. |
Send message Joined: 16 Dec 13 Posts: 23 Credit: 48,844,778 RAC: 35 ![]() ![]() ![]() |
I know it's probably not the solution you're looking for, but it's always a good idea to have a couple backup projects to get you through the times when tasks are scarce, whether that's due to high traffic, server problems, etc... Click here to see My Detailed BOINC Stats ![]() |
![]() Send message Joined: 18 Jul 09 Posts: 300 Credit: 303,565,482 RAC: 78 ![]() ![]() ![]() |
Let the Bitcoin Utopis run run down with out getting any new WUs from it. The MW units should increase then to a maximum of 40. |
![]() Send message Joined: 4 Jul 14 Posts: 8 Credit: 64,068,552 RAC: 0 ![]() ![]() |
The Bitcoin Utopia WUs aren't utilizing CPU or GPU. They're using miners attached to the computers. I don't see how they would affect anything. Also, the primary machines I'm concerned about are the servers. They aren't running any other project right now. They still won't pull down more than 24 WUs, and that usually involves some fighting. |
Send message Joined: 15 Aug 14 Posts: 6 Credit: 1,664,288 RAC: 87 ![]() ![]() |
The project has a very low maximum for CPU's of just 3 units. So if your server has 8 CPU's 24 units is the max limit. |
![]() Send message Joined: 4 Jul 14 Posts: 8 Credit: 64,068,552 RAC: 0 ![]() ![]() |
The project has a very low maximum for CPU's of just 3 units. I had to double-check all the machines to confirm, but the evidence definitely supports this. I can have dozens of WUs for a single GPU, but only 3 per CPU. That is a bit frustrating. Well, I thank you for at least shining some light on the issue at hand, but it is apparently an issue at a much higher level than I can fix. |
©2024 Astroinformatics Group