• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

NFS@Home app_config

  • Thread starter Deleted member 88227
  • Start date
D

Deleted member 88227

Guest
Anyone got an app_config for this project? I want to limit this project, and only this project, to a set number of CPU cores. Figure using an app_config would be the easiest way to do this.
 
I give up. This shit is so stupid.

NFS@Home was consuming all 20 threads of my 10c20t Intel rig. While two Moo! Wrapper GPU tasks were sitting idle (2 GPUs in this system).

So I made a couple config files for nfs@home. Used all three of these combinations.

<max_concurrent>18
<project_max_concurrent>18
<avg_ncpus>18

Which one of the combinations sort of worked... it only let NFS run 18 tasks... but the GPU tasks never got started. So I set no new work for NFS, aborted all the tasks and now the rig is just sitting idle with 2 idle Moo! Wrapper tasks sitting there doing nothing at all.

What. The. Actual. Fuck?
 
Is Moo! wrapper suspended? What does the event log say? Do you have the GPU suspended in any way? Double check the Activity > Use GPU's based on preferences and Options > computing preferences to see if you have GPU suspended when host is active.
 
The formula for limiting the total work units for a project is:

<app_config>
<project_max_concurrent>N</project_max_concurrent>
</app_config>

N being the total number of work units from the project you want running at any given time.

Also, if you are running the project but want to limit the total number of a specific app you would use the following I think:
<app_config>
<app>
<name>lasieve5f</name>
<max_concurrent>N</max_concurrent>
</app>
</app_config>

I think each app is listed below. They can be found in the client_state.xml file.
lasieved
lasievee
lasievef
lasieve5f

Each app uses different amounts of RAM. Change N to the number you want running at any given time.

So an example may look like this:

<app_config>
<app>
<name>lasievef</name>
<max_concurrent>4</max_concurrent>
</app>
<app>
<name>lasievee</name>
<max_concurrent>1</max_concurrent>
</app>
</app_config>
 
Last edited:
I did that. While it did limit it to the number I wanted (18) it did not let Moo! Wrapper run. They are still "ready to start". They're set to run always (the GPUs) Event log doesn't say much, but when I try to get WUs from other GPU projects (such as collatz) it says it doesn't need them.
 
I rebooted the whole rig already. But interesting enough I had the network work queue set to .01 so it would only download tasks as it completed them. I set it to 1 day and it just downloaded 200+ Moo! Wrapper tasks and 100+ pogs tasks. pogs is running on all 20 threads. Moo! Wrapper isn't running at all.
 
And I don't think it is the app_config preventing Moo!. And if you aborted all of the CPU work, then it isn't NFS. Do you have any other CPU work downloaded?
 
Can you confirm that they are GPU work units and not CPU Moo! work units?
 
No other CPU work is downloaded. Next to the Moo! Wrapper units it says (0.2C + 2NV) next to all of them.
 
I see that 1 of your rigs has both GPU and CPU work and 2 only has GPU. .... this is strange
 
Only one rig should have both, the HTPC. The Skillz2 is the host I am having issues with.
 
I suspended all projects. Not Moo! Wrapper Nothing.
I suspended Moo! Wrapper (plus all project). Nothing. Oddly enough it didn't even show up as being Suspended next to the project like pogs does.
Reset the project waiting to see what happens now.
 
Okay. I detached the project. After suspending everything I finally got Collatz to run on it's own. Going to reattach the project and see what happens.
 
Well damn I got it working. I had to use BAM! to detach the project and reattach it. Now she's chugging along like she's supposed to. 19 pogs tasks and 1 MW task running on 2 NV GPUs.
 
Let me know if you think it is better to run the tasks across both cards or running 1 per card. Also, let me know if you do the process phoneicis pointed out for running the different OpenCL application.
 
I was trying to test that out earlier, but it wasn't working. I've got two single 1080s in two rigs and 2x 980Ti in that rig. I want to see how well they do running two cards per task. I have not used any kind of tweaks for MW yet.

I will do some minor testing until I leave. Assuming I have time.
 
I would suggest sticking to saying Moo! because others may think you are referring to Milkyway when you abreviate as MW.
 
You must log in or register to reply here.
Back
Top