Message boards :
Questions and problems :
Work fetch problem with more than 1 ATI GPU, app_info & exclusions
Message board moderation
Previous · 1 · 2 · 3
Author | Message |
---|---|
Send message Joined: 6 Jul 10 Posts: 585 |
Seeing from the Rom reply that it's a question of updating the documentation to [<type>nvidia|ati|intel_gpu</type>] Coelum Non Animum Mutant, Qui Trans Mare Currunt |
Send message Joined: 16 Aug 12 Posts: 39 |
I had high hopes for 7.0.45. Unfortunately work fetch has not improved and the exclusions problem has not been helped at all. You will have to email the BOINC Alpha email list about that, as else the developers will not know about it. Discussed at length and simulations run. I think David may have fixed the problem but we'll have to see when the next BOINC version (7.0.53?) is released. Thanks David, Ageless, Erik, Jacob, John and all! Regards/Beyond |
Send message Joined: 6 Jul 10 Posts: 585 |
Seeing from the Rom reply that it's a question of updating the documentation to Since when I made this post and because many don't do the famous RTFM, it came out that nvidia and ati has to be capitalized in the type line. The wiki was corrected [with the question by the dev., why this could have been wrongly documented for so long], of course others asking then why it's case sensitive... well, at least now we know why there was loads of headscratching and unexplained behavior ;D. The option now [should] read [with one of the 3 generally referenced GPU brands] <type>NVIDIA|ATI|intel_gpu</type> Coelum Non Animum Mutant, Qui Trans Mare Currunt |
Send message Joined: 9 Nov 10 Posts: 63 |
Note: Because ROM spun a 7.0.53 build for Android several days ago (as indicated by http://boinc.berkeley.edu/dl/?C=M;O=D)... I'm expecting 7.0.54 to be the first build that has a potential fix for this thread's main issue. I'm seeing other oddities in the simulator, though, that indicate another fix may be necessary. I've posted them to the alpha list. Regards, Jacob |
Send message Joined: 16 Aug 12 Posts: 39 |
Since when I made this post and because many don't do the famous RTFM, it came out that nvidia and ati has to be capitalized in the type line. The wiki was corrected [with the question by the dev., why this could have been wrongly documented for so long], of course others asking then why it's case sensitive... well, at least now we know why there was loads of headscratching and unexplained behavior ;D. The option now [should] read [with one of the 3 generally referenced GPU brands] This bug has nothing to do with capitalization or syntax. I'd appreciate keeping this thread on the original topic: Work fetch problem with more than 1 ATI GPU, app_info & exclusions There's a proposed fix in the pipeline, maybe it will be in the next released version. Thanks Much, Beyond |
Send message Joined: 6 Jul 10 Posts: 585 |
Is this comment from the same as the Ed A on the alpha mail list? My post was a correction, in case a reader flips through with search and uses the wrong code! Thanks though for pointing out what the topic originally was ;>) Coelum Non Animum Mutant, Qui Trans Mare Currunt |
Send message Joined: 16 Aug 12 Posts: 39 |
Is this comment from the same as the Ed A on the alpha mail list? My post was a correction, in case a reader flips through with search and uses the wrong code! Thanks though for pointing out what the topic originally was ;>) Yep, just trying to keep the thread I started on topic and not going off in various directions. This bug has been percolating for a long time. Hopefully the next build will squash it :-) Regards/Ed |
Send message Joined: 9 Nov 10 Posts: 63 |
So, Ed... How's your work fetch been going lately? Did 7.0.54+ or 7.0.56+ fix your problem(s)? |
Send message Joined: 17 Aug 12 Posts: 11 |
well this thread inadvertently got locked 138 days ago for having reached the 90-day age limit - before the work fetch bug that is the topic of this thread was even solved. while some folks are already quite aware of it, i'm sure many others are not aware that the work fetch bug has in fact been resolved...that is to say, work buffers are now being maintained when crunching different projects on two or more GPUs of the same type (brand) in the same machine. because this information was absent from the very thread that originally addressed the problem, moderator Ageless was kind enough to temporarily unlock it so that i could post it. now i'm not entirely sure which revision of BOINC actually fixed the work fetch bug, but i do know that the fix was implemented into a BOINC revision released between the releases of v7.0.64 and v7.2.28...of course the fix very well may have initially been implemented in an unofficial (beta) version, and so 7.2.28 may very well be the first official (non-beta) release that contains the fix, but don't quote me on that. it has been suggested to me that BOINC v7.1.15 beta was the first release to contain the fix, but i have not confirmed it, nor have i ever used that particular BOINC release anyways. at any rate, i thought this information might come in handy to those who aren't running a fairly current version of BOINC and thus could still theoretically encounter the work fetch bug. thanks to Ageless for temporarily unlocking the thread long enough to post that the bug has been resolved, and to forum member Beyond for noticing the bug, alerting everyone to it, and devoting so much time and resources to testing for the fix. Eric |
Send message Joined: 14 Dec 13 Posts: 3 |
Since I started using 7.2.33, I can't contact the server. I get the following: 12/13/2013 5:36:04 PM | SETI@home | Fetching scheduler list 12/13/2013 5:36:05 PM | SETI@home | Master file download succeeded 12/13/2013 5:36:11 PM | SETI@home | Sending scheduler request: To fetch work. 12/13/2013 5:36:11 PM | SETI@home | Requesting new tasks for CPU and ATI 12/13/2013 5:36:12 PM | SETI@home | Scheduler request completed: got 0 new tasks 12/13/2013 5:36:12 PM | SETI@home | Server error: feeder not running 12/13/2013 6:19:26 PM | SETI@home | Sending scheduler request: To fetch work. 12/13/2013 6:19:26 PM | SETI@home | Requesting new tasks for CPU and ATI 12/13/2013 6:19:27 PM | SETI@home | Scheduler request completed: got 0 new tasks 12/13/2013 6:19:27 PM | SETI@home | Server error: feeder not running 12/13/2013 7:19:30 PM | SETI@home | Sending scheduler request: To fetch work. 12/13/2013 7:19:30 PM | SETI@home | Requesting new tasks for CPU and ATI 12/13/2013 7:19:32 PM | SETI@home | Scheduler request completed: got 0 new tasks 12/13/2013 7:19:32 PM | SETI@home | Server error: feeder not running It is probably something simple that you experienced users know about, but I need a little help, please. |
Send message Joined: 23 Apr 07 Posts: 1112 |
Since I started using 7.2.33, I can't contact the server. I get the following: What's that to do with this thread? This thread is about a work fetch problem when using an app_info and GPU exclusions, I think it's also been fixed. You also should ask that at the project in question, this is the Boinc Message Boards, it isn't Seti. The answer is the feeder isn't running, so they can't supply work, since fixed. Claggy |
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.