Message boards : Questions and problems : Boinc client run all program , why?
Message board moderation
Author | Message |
---|---|
Send message Joined: 8 Dec 15 Posts: 20 |
Hy all. Configuration problem :( I would like to: 1,download 1 project. 2,computing 3, upload data. 1 2 3 But current! : download 10 project , and computing all! AWWWW Hown to configure , one download and computing ? Thanks! |
Send message Joined: 29 Aug 05 Posts: 15581 |
Depends on the terminology what you mean with project. We speak of a project when we mean Seti@Home, Einstein@Home, Primegrid@Home, Collatz Conjecture, Virtual LHC, etc. The work that they run are tasks. So what do you mean with projects, the actual projects or the tasks from those projects? And what do you mean with 'download 10, computing all'? It depends on your hardware if that's possible. An i7 octo core can run 16 tasks at the same time, when hyperthreading is on. An i7 octo core with 2 GPUs can also run 10 tasks, even with hyperthreading off. It's difficult to give you an answer without some basic information about your system. |
Send message Joined: 8 Dec 15 Posts: 20 |
hownto configurate server? this render project. |
Send message Joined: 29 Aug 05 Posts: 15581 |
Okay, I take it you run a project yourself and you want all clients to run only one task, no matter how many CPUs they have? In Project Options, http://boinc.berkeley.edu/trac/wiki/ProjectOptions#Joblimits you can set Job Limits. <one_result_per_user_per_wu/>If set, send at most one instance of a given job to a given user. This increases the effectiveness of replication-based validation by making it more difficult for hackers to get all the instances of a given job. <one_result_per_host_per_wu/>If present, send at most one result of a given workunit to a given host. This is weaker than one_result_per_user_per_wu; it's useful if you're using homogeneous redundancy and most of the hosts of a particular class belong to a single user. See also http://boinc.berkeley.edu/trac/wiki/ProjectOptions#Joblimitsadvanced for more advanced job limits, and for instance the capability of setting per processor how many tasks it can run. You set these in the config.xml file on the server. |
Send message Joined: 8 Dec 15 Posts: 20 |
Thanks Ageless! rendering app :) |
Send message Joined: 5 Oct 06 Posts: 5137 |
Is the rendering app multi-threaded (uses all cores on a multi-core CPU)? If so, it should be deployed on the server with an 'MT' plan_class See https://boinc.berkeley.edu/trac/wiki/AppPlanSpec, where the keyword is still 'multicore' instead of 'multithread'. mt is available as a Predefined plan class |
Send message Joined: 8 Dec 15 Posts: 20 |
awww.... wrapper_26016_windows_x86_64.exe <job_desc> <task> <application>tobb_21769.exe</application> |
Send message Joined: 29 Aug 05 Posts: 15581 |
awww.... Without you saying what you did, we cannot help. Give an example of what you added for job restrictions in your config.xml |
Send message Joined: 5 Oct 06 Posts: 5137 |
I think we are being held back from advising you fully because of the language barrier. It might be better for you to ask for help in your native language. The SZTAKI Institute in Budapest has a lot of experience in hosting BOINC servers and running BOINC projects: it hosted the most recent BOINC Workshop last year, and apart from Prof. Peter Kacsuk, Jozsef Kovacs, Robert Lovas, Attila Marosii, and Adam Visegradi also contributed. I'm sure that somebody from that list will be able to help you more effectively than we can, or at least put you in touch with somebody who can. |
Send message Joined: 8 Dec 15 Posts: 20 |
config.xml <?xml version="1.0" ?> <boinc> <config> <daily_result_quota>50000</daily_result_quota> <one_result_per_host_per_wu/> <max_wus_to_send>1000</max_wus_to_send> <!-- <max_wus_in_progress> 1 </max_wus_in_progress> <max_wus_in_progress_gpu> 1 </max_wus_in_progress_gpu> --> |
Send message Joined: 8 Dec 15 Posts: 20 |
OK. thanks |
Send message Joined: 20 Nov 12 Posts: 801 |
<?xml version="1.0" ?> I'm not entirely sure what BOINC's not-really-XML parser will think of this. You may want to remove it. <!-- This starts a comment in XML. --> And this ends the comment. As a result <max_wus_in_progress> and <max_wus_in_progress_gpu> are unlikely to be in effect. Try removing the comment markers. Did you remember to restart the scheduler after changing the config.xml? edit: In the screenshot I'm guessing you have 8 tasks running. Do you have 8-core CPU? <max_wus_in_progress> limits the number of tasks per core. When you set max_wus_in_progress=1 the client will receive 1 task for every core it has. If you want to send really only one task you'd need to use advanced job limits and set up config_aux.xml . And, as Richard commented, if this is a multi-threaded application you should use the mt plan class. |
Copyright © 2025 University of California.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License,
Version 1.2 or any later version published by the Free Software Foundation.