| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Conquer Cancer
Thread: No tasks are available... |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 14
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I was checking the activity for HCC this morning and I saw the number of work units is lower than what I expected. I went to the HCC forum to see if I could find a reason for this, but I did not find any. The closest I found is a thread for "Cache was dry this morning" which seems to be addressing a problem getting GPU work units and the last post was 11/07/2012. I did check my messages and the last ones sent are
----------------------------------------11/15/2012 9:10:29 AM World Community Grid Sending scheduler request: To fetch work. 11/15/2012 9:10:29 AM World Community Grid Requesting new tasks for GPU 11/15/2012 9:10:31 AM World Community Grid Scheduler request completed: got 0 new tasks 11/15/2012 9:10:31 AM World Community Grid Message from server: NVIDIA GPU: Upgrade to the latest driver to process tasks using your computer's GPU 11/15/2012 9:10:31 AM World Community Grid Message from server: No tasks sent 11/15/2012 9:10:31 AM World Community Grid Message from server: No tasks are available for Help Conquer Cancer 11/15/2012 9:10:31 AM World Community Grid Message from server: The Help Conquer Cancer NVIDIA GPU application cannot run on your graphics card [Edit 1 times, last edit by Former Member at Nov 15, 2012 5:35:10 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Note that my computer is a lenovo ThinkPad and I do not intend to use the GPU for HCC work. I have another computer which has a graphics card which does allow the GPU work.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Move the device to a different profile [there is a possible of 4], and deselect the GPU function in there. If the device shares a profile with another device that can GPU crunch, then it will try to fetch those, and the server then checking if the device actually can.
----------------------------------------Alternately, on the device add this code <no_gpus>1</no_gpus> to the cc_config.xml <options> section and that client wont ever try to get GPU tasks. [Edit 1 times, last edit by Former Member at Nov 15, 2012 5:53:46 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thanks Rob. I am using a different profile for this computer now and I am not getting the server messages about the GPU. I will need to check two other computers to see if I have the same problem.
I am still not getting any new work for the ThinkPad. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Suppose that device was moved to a profile that also has HCC selected, but to not send GPU work. The CPU supply is intermittent at times, so what is your message log now saying?
----------------------------------------Getting variations of a warning there's no DDDT2 or No HCC+DDDT2 on my laptop, or there is but only for other platforms... don't know why that is, but after a while the HCC for CPU trickle in again. [Edit 1 times, last edit by Former Member at Nov 15, 2012 6:40:15 PM] |
||
|
|
joeperry39@gmail.com
Advanced Cruncher USA Joined: Nov 22, 2006 Post Count: 140 Status: Offline Project Badges: 
|
I also noticed that the number of wu's waiting to process was far less than normally displayed. However, when I looked at the "estimated time to completion," I discovered that the required time had doubled for each job on the list ( 2 computers - same problem with both and both running HCC 7.05)
----------------------------------------Perhaps there is a change in the size of the WU's???  "Everything in moderation, including moderation" -- Mark Twain |
||
|
|
Bearcat
Master Cruncher USA Joined: Jan 6, 2007 Post Count: 2803 Status: Offline Project Badges: 
|
In your profile settings, make sure the process to allow GPU wu's is unchecked. Should stop the message about your Nvidia GPU.
----------------------------------------
Crunching for humanity since 2007!
 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello osugrad,
----------------------------------------I also noticed that the number of wu's waiting to process was far less than normally displayed. However, when I looked at the "estimated time to completion," I discovered that the required time had doubled for each job on the list ( 2 computers - same problem with both and both running HCC 7.05) Perhaps there is a change in the size of the WU's??? Yes. We started out with only 1 image per work unit. HCC 7.05 now has 2 images per work unit, with a check point between the images, so a work unit takes twice as long to run. This reduces the work load on the server.  Now we seem to be having trouble validating the old 6.56 GPU work units.Lawrence [Edit 1 times, last edit by Former Member at Nov 16, 2012 10:33:01 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have reviewed the WUs I am getting and the message log, and I now find that I am getting WUs one at a time as opposed to the multiple WUs I would see downloaded at a previous time. I am also seeing a CPU time ranging from about 2.5 hours to 4 hours, while previously the time was 1.0 to 1.5 hours (take that 1.0 as about 55 minutes).
I also investigated the tasks in PV which I have over 15 pages (228 WUs) and a good number have wingmen with a Status of No Reply. For the computer with a usable GPU, these WUs have 3 to 4 wingmen with the No Reply (or Error) status. So, for now anyways, things look to be fairly normal. |
||
|
|
joeperry39@gmail.com
Advanced Cruncher USA Joined: Nov 22, 2006 Post Count: 140 Status: Offline Project Badges:
|
Yes. We started out with only 1 image per work unit. HCC 7.05 now has 2 images per work unit, with a check point between the images, so a work unit takes twice as long to run. This reduces the work load on the server. That would explain the increase in time -- Lawrence, thanks for the reply. Now we seem to be having trouble validating the old 6.56 GPU work units.Ah, the ''Law of Unintended Consequences'' strikes again  "Everything in moderation, including moderation" -- Mark Twain |
||
|
|
|