| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: question about BOINC error messages |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 5
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hiya,
I joined the WCG only a day or so ago. I'm already using BOINC for the ClimatePrediction programme. While trying to get the the WCG configured I selected all available projects, and while I thought the WCG was computing tasks ok, now I'm starting to get red error messages like this: 14/06/2009 07:33:10 World Community Grid Scheduler request completed: got 0 new tasks 14/06/2009 07:33:10 World Community Grid Message from server: No work sent 14/06/2009 07:33:10 World Community Grid Message from server: (won't finish in time) Computer on 59.7% of time, BOINC on 51.3% of that This is just a sample of several consecutive messages, I have no idea what they mean. And now I find that although the WCG project is attached, the actual tasks have disappeared and all I see are the Climate Prediction tasks. I know that if there is no available work for a project, no work will be sent. But I think this has more to do with how I share out resources between projects on my computer...am I right? Any help would be appreciated. Tim |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Sorry, but i think with the CPDN tasks on board, several i understand and the time the device is on, the client thinks it wont be able to finish them in time, so it stops all work fetching.
----------------------------------------Message from server: (won't finish in time) Computer on 59.7% of time, BOINC on 51.3% of that Leave the computer on longer and the problem may resolve itself. How many hours do the CPDN jobs need each and what is their deadline? When you look in the tasks tab of BOINC Manager (advanced view), are they running high priority? ADDED: The way I managed CPDN, I did long running of 1600 hours taking months is, to fetch a single job on a multicore device, then set for this project the work fetch to NO. This way, CPDN will take just 1 core, will probably use much if all the time of 1 core and other cores can work on the additional projects. When CPDN is finished, fetch a new job, and set the option to NO work fetch again.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 2 times, last edit by Sekerob at Jun 14, 2009 12:35:53 PM] |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 3010 Status: Offline Project Badges: 
|
And now I find that although the WCG project is attached, the actual tasks have disappeared They've probably "disappeared" from your BOINC session due to them being completed - although, if you look on the My Grid | Results Page, you'll probably find them there (well, for a short time anyway - they will eventually 'drop off' from there, once they get sent back to the respective project scientists)  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
How many hours do the CPDN jobs need each and what is their deadline? When you look in the tasks tab of BOINC Manager (advanced view), are they running high priority? I have two CPDN tasks with another one waiting. One task's elapsed time is 446 hrs 52 mins; 161 hrs to completion, and its deadline is 15/01/2010. The other one's elapsed time is 378hrs 47mins; 90hrs to completion, and a deadline of 04/08/2009. The tasks just say they're running, not set to 'high priority'. I leave my PC on for days unattended, especially overnight. I set tasks to run all the time when it's idle. I don't always like them running when I'm working on the PC because I often use CPU-intensive software and I need as much power as possible for it. ADDED: The way I managed CPDN, I did long running of 1600 hours taking months is, to fetch a single job on a multicore device, then set for this project the work fetch to NO. This way, CPDN will take just 1 core, will probably use much if all the time of 1 core and other cores can work on the additional projects. When CPDN is finished, fetch a new job, and set the option to NO work fetch again. If I understand right, you're saying that because I'm adding new tasks and diverting CPU cycles from CPDN, it's estimating that it won't meet the deadline for them? I'll try your suggestion. Thanks for the input. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Sounds like CPDN work units are monsterous, taking several hundred hours to complete but come with a very long deadline, many months. WCG work units only take a few hous but have deadlines much shorter, around 10 days. So your BOINC client thinks it's overcommitted with CPDN so it doesn't download anything from WCG because it can't return them in time.
The only solution I see is to run 1 project at a time. Meaning, crunch the CPDN tasks and return them, then detach from that project and attach to WCG. Not exactly convenient but it's the best I can come up with for now. |
||
|
|
|