Message boards :
Questions and problems :
balance un my boinc through app_config and not fully getting as much work as it should
Message board moderation
Author | Message |
---|---|
Send message Joined: 7 May 15 Posts: 23 |
So I kinda like to keep some balance so i'll explain my configs as they stand first: I have 2 gpus and a 4 core cpu I dedicate GPU 0 to einstein I dedicate GPU 1 (fastest) to SETI@home My CPU Based projects have set to max concurrent as 1 I wish to actively work on all projects as much as possible versus letting boinc decide one day to work on those and another day to work on the other ones at its own decision Right now I have run into an issue and its not using all the cores yet work units are available to the other projects. GPU 0 running einstein with 1 cpu core GPU 1 runnign seti@home ATLAS is using one cpu core LHC us running one cpu core for each of the 2 subproject tasks its running that totals four cpu cores at the moment but sometimes I go a day without using one of my cpu cores and yet projects liek enigma have work units available. Noticed it does say cpu not highest priority is there anyway to have to get work for spare cores when it comes down to it and not just leave cores unused? |
Send message Joined: 20 Nov 12 Posts: 801 |
At the moment <max_concurrent> isn't considered in work fetch. Work fetch thinks you have enough tasks in cache but doesn't realise not all of them can't be run yet. Tweaking cache size to smaller or larger or lowering <rec_half_life_days> might help. |
Copyright © 2024 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.