Credits per day drop to 1/3

Message boards : Questions and problems : Credits per day drop to 1/3
Message board moderation

To post messages, you must log in.

AuthorMessage
liderbug

Send message
Joined: 10 Dec 11
Posts: 24
United States
Message 69932 - Posted: 3 Jun 2016, 15:01:43 UTC

Last week I was running FC19 with a version of Boinc from about 3 years ago. My CPU is a AMD 6 core and I was averaging 30,000 credits per day. Last week I upgraded to FC22 and reloaded Boinc (DNF) My credits dropped to < 10,000. Any thoughts?
Thanks
ID: 69932 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15477
Netherlands
Message 69933 - Posted: 3 Jun 2016, 15:42:24 UTC - in response to Message 69932.  

BOINC doesn't do any science calculations, that's up to the project. Giving out credit is also something the projects do, so you best ask this question at the project you see this at. Perhaps that they've got a new application that runs slower.
ID: 69933 · Report as offensive
SekeRob2

Send message
Joined: 6 Jul 10
Posts: 585
Italy
Message 69943 - Posted: 3 Jun 2016, 18:17:38 UTC - in response to Message 69933.  
Last modified: 3 Jun 2016, 18:18:44 UTC

Does the project know something called Reliable rating status and Zero Redundant? If the client was not recognized as the old one, those statuses at WCG at least are lost, so lots of new work goes first via the Pending Validation status and needs a wingman to verify... this causes lots of delay and so the initial big drop for the day, but recouped in the future.
Coelum Non Animum Mutant, Qui Trans Mare Currunt
ID: 69943 · Report as offensive
liderbug

Send message
Joined: 10 Dec 11
Posts: 24
United States
Message 69957 - Posted: 4 Jun 2016, 14:22:25 UTC

More detail: I backed up my home dir. De-raided my system (kept getting failing drive - another story). Loaded FC22 from scratch - dnf install(ed) boinc-client boincmgr. When I started the first time I said Connect to BAM I'm returning... it knew who I was and put everything back and started running at 33%.


PS - I've found I can ONLY start boinc or mgr after I "cd /var/lib/boinc" - if not nothing happens.
ID: 69957 · Report as offensive
liderbug

Send message
Joined: 10 Dec 11
Posts: 24
United States
Message 70643 - Posted: 2 Jul 2016, 15:25:39 UTC - in response to Message 69957.  

After much time on the web searching for a solution to my problem I have come the conclusion that under FC-19 and the Boinc/projects of the load that my GPU was being used. Now it's not. What determines if the GPU is to be used. I have to think, Boinc, because the MGR log says "No usable gpu found" comes from the boingmgr log. My card is a GeForce GTS 250, but not found in the usable list - by Boinc.
ID: 70643 · Report as offensive
SekeRob2

Send message
Joined: 6 Jul 10
Posts: 585
Italy
Message 70675 - Posted: 4 Jul 2016, 8:09:23 UTC - in response to Message 70643.  

"What determines if the GPU is to be used."

Many a reason, but "No usable gpu found", hints at the good old race condition... the boinc daemon is loaded -before- the display drivers have been fully initiated. Try a boinc-client restart to see if it then will see the GPGPU.
Coelum Non Animum Mutant, Qui Trans Mare Currunt
ID: 70675 · Report as offensive
liderbug

Send message
Joined: 10 Dec 11
Posts: 24
United States
Message 70738 - Posted: 6 Jul 2016, 22:32:49 UTC

"... to use CUDA install xorg-x11-drv-nvidia-libs..."
"Not found"
search... rpm ..Not found WTF? go get it manually
Fedora 15, 16, 17, 18, 19, 20 ah, 21? 22?
Oh we decided it was out-dated so it's no longer supported.

To put it mildly - [redacted]
ID: 70738 · Report as offensive
liderbug

Send message
Joined: 10 Dec 11
Posts: 24
United States
Message 70754 - Posted: 8 Jul 2016, 13:39:29 UTC

SOLVED. (good news)

(the bad) Not sure how I did it... but I have a Seti ... CUDA(60).
I downloaded
NVIDIA-Linux-x86_64-340.96.run
xorg-x11-drv-nvidia-libs-346...rpm
dnf install cuda
dnf install kernel-devel kernel-headers
dnf install libvdpau
dnf install xorg-x11-drv-nvidia
dnf install kmod kmod-libs kmod-devel libmikmod libmikmod-devel mikmod
dnf install $(nvidia-detect)
dnf install vdpauinfo libva-vdpau-driver libva-utils
dnf install plymouth-theme-solar

and somehow I got/get:
CUDA: NVIDIA GPU 0: GeForce GTS 250 (driver version 340.96, CUDA version 6.5, compute capability 1.1, 1023MB, 943MB available, 622 GFLOPS peak)
07-Jul-2016 07:02:29 [---] OpenCL: NVIDIA GPU 0: GeForce GTS 250 (driver version 340.96, device version OpenCL 1.0 CUDA, 1023MB, 943MB available, 622 GFLOPS peak)
07-Jul-2016 07:02:29 [---] [coproc] NVIDIA library reports 1 GPU

So it looks like I may be back - time will tell.
ID: 70754 · Report as offensive

Message boards : Questions and problems : Credits per day drop to 1/3

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.