One computer is getting no work anymore

Mike
Mike
Joined: 22 Feb 05
Posts: 5
Credit: 168,660,556
RAC: 134,232
Topic 219851

I have 4 computers running this application. One of them has not received and work for quite a long time. I don't know why. Tried resetting, deleting and reinstalling application. Nothing.

Zalster
Zalster
Joined: 26 Nov 13
Posts: 2,992
Credit: 3,309,026,692
RAC: 707,863

Check your OpenCl and make

Check your OpenCl and make sure that Windows didn't install their version of OpenCl.  Otherwise download the nvidia driver (before 435) and do a clean install.  This is in log report

 

 

2019-10-27 17:27:32.4135 [PID=3597 ]    [version] Checking plan class 'FGRPopencl1K-nvidia'
2019-10-27 17:27:32.4135 [PID=3597 ]    [version] parsed project prefs setting 'gpu_util_fgrp': 1.000000
2019-10-27 17:27:32.4135 [PID=3597 ]    [version] NVidia device (or driver) doesn't support OpenCL
2019-10-27 17:27:32.4135 [PID=3597 ]    [version] Checking plan class 'FGRPopenclTV-nvidia'
2019-10-27 17:27:32.4135 [PID=3597 ]    [version] parsed project prefs setting 'gpu_util_fgrp': 1.000000
2019-10-27 17:27:32.4136 [PID=3597 ]    [version] NVidia device (or driver) doesn't support OpenCL

Mike
Mike
Joined: 22 Feb 05
Posts: 5
Credit: 168,660,556
RAC: 134,232

NVIDIA driver is old and has

NVIDIA driver is old and has been running fine. Asteroids APP using it perfectly right now. Too many OpenCL files in my C: drive to know which one is being used, but some are in the Nvidia folder and old.

 

No luck with just this one machine.

 

Richie
Richie
Joined: 7 Mar 14
Posts: 519
Credit: 1,609,369,422
RAC: 797,921

Asteroids GPU tasks use Cuda,

Asteroids GPU tasks use Cuda, not OpenCL.

http://asteroidsathome.net/boinc/apps.php

Your system currently doesn't have proper OpenCL capability installed, like Zalster pointed out. You need to reinstall Nvidia driver. Download GPU driver from Nvidia, then disconnect internet connection, run Nvidia installer, choose custom installation and choose clean install. Then reboot and connect internet.

DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1,287
Credit: 1,490,744,547
RAC: 987,775

Stop boinc when you do the

Stop boinc when you do the install or you'll probably error out whatever WU you have running at the time.

Dp
Dp
Joined: 27 Aug 05
Posts: 12
Credit: 66,725,575
RAC: 1,458

I have a similar

I have a similar problem.

Two AMD cards running other BOINC projects with no problem. It stopped E@home aprox. 2 weeks ago and has yet to dl any tasks.

Any suggestions?

TIA

 RWP

mikey
mikey
Joined: 22 Jan 05
Posts: 5,338
Credit: 526,437,303
RAC: 49,273

Dp wrote:I have a similar

Dp wrote:

I have a similar problem.

Two AMD cards running other BOINC projects with no problem. It stopped E@home aprox. 2 weeks ago and has yet to dl any tasks.

Any suggestions?

TIA

 RWP

When you look in the log  what does it say when you click to update the project?

Dp
Dp
Joined: 27 Aug 05
Posts: 12
Credit: 66,725,575
RAC: 1,458

This my event log 11/2/2019

This my event log

11/2/2019 3:20:52 PM | Milkyway@Home | Starting task de_modfit_14_bundle4_testing_3s4f_2_1570151403_7911598_3
11/2/2019 3:20:54 PM | Einstein@Home | Sending scheduler request: Requested by user.
11/2/2019 3:20:54 PM | Einstein@Home | Not requesting tasks: don't need (CPU: job cache full; AMD/ATI GPU: not highest priority project)
11/2/2019 3:20:56 PM | Einstein@Home | Scheduler request completed
11/2/2019 3:21:45 PM | Milkyway@Home | Computation for task de_modfit_81_bundle4_4s_south4s_bgset_2_1570151403_14427702_1 finished

 

Thanks for getting back

RWP

Holmis
Joined: 4 Jan 05
Posts: 1,037
Credit: 729,365,387
RAC: 195,972

Dp wrote:11/2/2019 3:20:54 PM

Dp wrote:
11/2/2019 3:20:54 PM | Einstein@Home | Not requesting tasks: don't need (CPU: job cache full; AMD/ATI GPU: not highest priority project)

This line tells you that Boinc thinks the CPU cache is full and that Einstein has run enough GPU tasks recently and now it's the other projects turn. When your other projects have got to run GPU tasks for a while their debt will be lowered and then Boinc will download more work from Einstein. Just be patient and let thing run, it looks like it's working as intended.

archae86
archae86
Joined: 6 Dec 05
Posts: 2,705
Credit: 2,565,994,546
RAC: 3,485,698

The evenness of work fetch

The evenness of work fetch and execution gets thrown off by periods when projects cannot be reached or don't have work.

Setting a large work cache size (multiple days for the sum of "store at least" and "store up to an additional") means the side effect of over-fetching available work when more highly desired work is not available can last quite a while.

Possibly you might like the project alternation behavior better if you set a very small work cache size.

Dp
Dp
Joined: 27 Aug 05
Posts: 12
Credit: 66,725,575
RAC: 1,458

Thanks to all I'll be

Thanks to all

I'll be patient

RWP

 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.