Issue with preferences
13 Mar 2017 4:48:13 UTC
Topic 206104
(moderation:
Today I changed my GTX660 to run 1 task at a time from the 2 it had been running. GPUZ showed me I should go back to 2. I changed the factor back to .5 from 1 and saved the change. It is still stuck running 1. It is as all the work I had cached, 1.5 days worth, was branded to run 1 at a time.
That's not an issue but
)
That's not an issue but regular behavior. New settings apply for new tasks only as has been mentioned for example in the MORE UNITS ON GPU thread. If you want it to apply immediately then you may just abort any old task and wait a minute for new tasks to be downloaded.
That is correct behavior.
)
That is correct behavior. Only tasks that are downloaded after the preference change are affected by the change. Tasks that are already on you computer are not affected. The Client does not know about the special Einstein@home specific preferences regarding GPU apps only the server does and adjusts settings on the fly when sending new tasks to your computer.
Christian Beer wrote:That is
)
And once new tasks with the new preference has been downloaded the preference will affect all tasks previously downloaded.
To speed up the process one can increase the cache setting so Boinc asks for new tasks and then reset it to whatever one wants.
Betreger wrote:...It is still
)
It's not stuck running 1 and all tasks in your cache can be run in pairs as long as a new task is downloaded after you changed to 0.5. Please don't abort anything. Just make a small increase in your work cache setting, just enough to trigger a work fetch of a new GPU task. try going from 1.5 to 1.6 days or whatever you need to get the client to ask for work. Once a new task is sent, your BOINC client will become aware of the change and will start a second GPU task. You can then restore the work cache setting to the former value. There will be two tasks crunching even though BOINC manager will show 1 GPU (instead of 0.5 GPUs) for each task until the old tasks are cleared.
EDIT: Sorry, Holmis beat me to it :-).
Cheers,
Gary.
Ah, that fixed it, you guys
)
Ah, that fixed it, you guys are smart.
Thanx
My problem was caused because
)
Betreger wrote: My problem
)
That is still correct. The project preferences are not Client-related, they are App-related and refer to the science applications of a project.
Christian Beer wrote:Betreger
)
Betreger wrote:Christian Beer
)
And that is not a valid point because a volunteer changing client-related preferences would be left with the incorrect impression that both an 'update' and a download of new work would be needed for the preference change to apply.
What could be done is to add the requirement to download new work to the warnings about adjusting the GPU utilization factor because that is the specific preference change to which this requirement applies.
Whilst thinking about this, I'm of the opinion that the 'dire consequences' warning should be modified to be a little less 'dire' and a little more informative about the whole process. The last 'CYA' sentence in the utilization factor warning should be the bit that is emphasised as the risk should always be with the user. We don't really need three identical warnings for the individual settings, just a single more detailed set of notes about all of them at the start. I would suggest something like this:-
Run CPU versions of applications for which GPU versions are available: Yes | No
Run multiple GPU tasks concurrently on sufficiently powerful GPUs
Please note that if you have any doubt about the suitability of your GPU, you should post a query in the forum to seek advice before changing any of the settings below. For multiple tasks, you need larger amounts of GPU RAM.
If you change settings (with advice or not), it is entirely at your own risk!
The running of multiple GPU tasks is likely to increase the loading, power consumption and operating temperature of your system and could cause system instability or even hardware damage, particularly if your cooling solution is not adequate for the increased load.
You should only change the factor for the particular type(s) of GPU task(s) that you wish to run on your computer(s). Not all types of tasks may be available at any particular point in time. The factor represents the 'fraction' of a GPU device that a single task will 'consume'. For example, if you wish to run two tasks concurrently per device, you would set the factor to 0.5. Default values are all 1.0 (run 1 task per device).
If you change any of these factors, it will only be applied by your BOINC client after it has downloaded a new task of that particular type. For the impatient, you could use a small increase in the work cache setting to trigger an immediate work fetch. The change will also apply to all previously downloaded work of that type on your computer.
GPU utilization factor of BRP (Binary Radio Pulsar) apps: 1
GPU utilization factor of FGRP (Fermi Gamma-Ray Pulsar) apps: 1
GPU utilization factor of GW (Gravitational Wave) apps: 1
.....
I realise that the amount of extra information might be viewed as excessive but I feel there needs to be a little more detail to help new/inexperienced volunteers to get started in the first place. It's only a suggestion - something the Devs might like to think about.
Cheers,
Gary.
Gary, this seems a bit
)
Gary, this seems a bit excessive but I agree three times the same warning is also a bit too much. Since this is also a more technical problem could you please open an issue in our Drupal issue tracker with that? We are going to discuss how to implement and the wording there. This is going to be moved to "Phase II" with all the other enhancements that the community wants to have. If it is not in the tracker it will soon be forgotten in the forums here.