[Support] Linuxserver.io - Folding@home


Recommended Posts

  • 1 month later...
  • 2 weeks later...

Im getting this strange error after starting up F@H after some time, "ERROR:WU00:FS00:Exception: Failed to remove directory 'work/00': Directory not empty"

I tried stopping the Docker and Deleting the F@H folder out of Appdata, but that didn't resolve the issue.


Meanwhile folding is still going on normally, and it works without issues.

Link to comment
  • 2 weeks later...

6 days ago I had GPU folding working great.  Today I see:

2:58:36:      GPU 0: Bus:1 Slot:0 Func:0 NVIDIA:7 TU116 [GeForce GTX 1660 SUPER]
02:58:36:       CUDA: Not detected: cuInit() returned 804
02:58:36:     OpenCL: Not detected: clGetPlatformIDs() returned -1001
 

so therefore my GPU folding is broke and flagged as disabled.  I deleted/reinstalled the docker (keeping and reusing the FAH appdata folders) and it made no difference.  So I am left wondering, did UNRAID NVIDIA stop working when they pulled the plug on it?  I didnt think that was supposed to happen?

 

If the only fix here is to install the unraid beta w/the new nvidia plug in,  then is the beta ok to install on top of nvidia unraid?

Edited by nealbscott
Link to comment

nealbscott, 

I had the same issue and discovered that the docker image had been updated. Something in the update definitely broke functionality for me. So I reverted to the previous docker image by changing the variable for the repository to the specific last build that worked and updated the container: linuxserver/foldingathome:7.6.21-ls21

Folding@Home as been running correctly on the GTX 970 card ever since.

Link to comment
  • 2 weeks later...
On 12/6/2020 at 11:04 AM, apotek said:

I had the same issue and discovered that the docker image had been updated. Something in the update definitely broke functionality for me. So I reverted to the previous docker image by changing the variable for the repository to the specific last build that worked and updated the container: linuxserver/foldingathome:7.6.21-ls21

Folding@Home as been running correctly on the GTX 970 card ever since.

apotek,

 

Thank you very much!  I follow exactly what you said in your post, restarted the F@H container, and now my GPU in my unRAID machine is folding again!

 

Very grateful!

 

One question ... if a person is installing Linuxserver's Folding at Home app on their unRAID machine for the first time, how can they get the GPU ID with out the unRAID nvidia plug in?

 

Thank you for all your help!!

Link to comment
On 12/19/2020 at 12:44 AM, apotek said:

As long as the nvidia driver is installed, open up the Unraid web terminal and type:  nvidia-smi -L

This will list the UUID.  Or you can instill the plugin GPU Statistics which has this info on the plugin's Settings page and adds some handy monitoring stats on the Dashboard page. All thanks to b3rs3rk for his hard work. 

 

apotek,

 

Thank you again!

Link to comment
  • 3 weeks later...

Hello,

During the night, the container has been automatically updated from 7.6.21-ls25 to 7.6.21-ls26. Since then, the existing GPU slot is disabled with the following message in the log :

08:22:15:WARNING:FS01:No CUDA or OpenCL 1.2+ support detected for GPU slot 01: gpu:43:0 GP106GL [Quadro P2000] [MED-XN71]  3935.  Disabling.

The server has been folding for at least 10 days, nothing else has changed in the setup (Unraid 6.9.0-beta35 with Nvidia Driver). Output of nvidia-smi :

Sat Jan  9 09:26:32 2021       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 455.45.01    Driver Version: 455.45.01    CUDA Version: 11.1     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  Quadro P2000        Off  | 00000000:2B:00.0 Off |                  N/A |
| 64%   35C    P0    16W /  75W |      0MiB /  5059MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+

So, I think this release has an issue, probably far beyond my technical skills ...

Does anybody know how, as a workaround before a fix, I could downgrade from 7.6.21-ls26 to 7.6.21-ls25 which was perfectly fine ?

 

Thanks in advance for the support.

 

Edit : I got support on linuxserver.io discord channel, problem temporarily solved. Thanks again !

The container has been updated with new nvidia binaries, and it seems it is no longer compatible with my installed drivers (455.45.01). I don't know if it would be OK with the regular drivers proposed by Nvidia Driver plugin, ie 455.38 in 6.9.0-beta35. For sure, the best solution would be to have the latest stable Nvidia drivers (v460.32.03) in Unraid, as the latest version of the container is fully compatible with them.

The workaround to downgrade to the previous version of the container is to edit the template repository from "linuxserver/foldingathome" to "linuxserver/foldingathome:7.6.21-ls25".

Folding again, which is the most important for the moment !

But this example clearly raises the problem of regular updates of Nvidia drivers by Limetech when 6.9.0 is stable ...

Until then, I'll stick to ls25 version of the container.

 

Edit 2 : As I had posted an issue on github, the developers proposed me to test a dev build. I did and it worked fine, so I suppose a new version of the container will be soon publicly available and others should not have the same issue. Thanks for the responsiveness to @aptalca and linuxserver team !

Edited by Gnomuz
  • Like 1
Link to comment
11 hours ago, Gnomuz said:

No need to use an incognito window for me, but I rarely use it anyway. I prefer FAHControl which gives you much more control and features. Maybe you can try and empty your browser cache ?

I'm trying to link it to FAH control but it's stuck on "updating", any ideas there?

 

edit: nevermind, i had a typo in the settings that I missed.

Edited by sage2050
Link to comment
  • 2 months later...

Hello!

 

I need some help getting a folding slot working again.

 

I have a GTX 690, which is a dual GPU.  I had Folding @ Home running fine on the CPU and one of the GPU's ... and then I made some changes to the docker settings to try and get the second GPU  to start folding also.

 

The setting I changed were to add a "Variable" -> Nvidia Visible Devices: (that I called GPU #2) ... & added the GPU ID ... and then pressed 'APPLY'.

 

The docker started back up fine ... but then my first GPU wasn't folding.  The CPU and the second GPU are folding.  When I open the web GUI it tells me the 'GPU disabled' ... when I mouse over the red dot for that GPU, it tells me 'Folding slot disabled'.  Does anyone know how to get both GPU's folding at the same time?

 

Thank you for all your help!!

F@H1.png

F@H2.png

Link to comment
  • 3 weeks later...
  • 2 months later...

My F@H docker image was running just fine... then after a reboot the container starts, but the web console times out (http://10.0.0.52:7396/)

 

The netdata plugin shows that nothing much is happening on the GPUs or CPU, so I guess that my F@H somehow got broken, and is pretty silent about why. Stopping and starting again doesn't change. GPU IDs remain the same as before the reboot. No errors about GPUs or drivers, and the GPUs are recognized.

 

Latest logs here: https://pastebin.com/g1vm41Qy

 

I've scanned this thread, and upgraded to the latest, but no change in behavior. I've looked for logs besides those produced directly by the F@H control client, under /var/log, but nothing interesting.

 

Posting here in the hope that one of you recognizes the pattern and can nudge me in the right direction.

 

Cheers.

 

Link to comment
  • 1 year later...
  • 2 months later...
Quote

08:06:01:      GPU 0: Bus:41 Slot:0 Func:0 NVIDIA:7 TU106 [Geforce RTX 2060]
08:06:01:       CUDA: Not detected: Failed to open dynamic library 'libcuda.so':
08:06:01:             libcuda.so: cannot open shared object file: No such file or
08:06:01:             directory
08:06:01:     OpenCL: Not detected: clGetPlatformIDs() returned -1001
...
08:06:01:FS00:Initialized folding slot 00: cpu:10
08:06:01:WARNING:FS01:No CUDA or OpenCL 1.2+ support detected for GPU slot 01: gpu:41:0 TU106 [Geforce RTX 2060].  Disabling.
08:06:01:WU00:FS00:Starting



I am seeing this when I try and use my GPU in Folding at home.

I included pics of the config for the docker and the way I set the Nvidia variables.
I am on driver 515.76
using Unraid 6.11.1

Let me know if I can provide any more info
Thanks

Screenshot 2022-10-08 021346.jpg

Screenshot 2022-10-08 021046.jpg

Screenshot 2022-10-08 020959.jpg

Link to comment
  • 3 months later...

 So, I'm very proud of myself. I am NOT a power user, but I was able to get F@H up and running on my Unraid box (Core i3 10100, rtx2060, 16gb ram + storage not relevant here). 

 

However, that CPU is rather pointless to be folding on. Just not enough horsepower. What do I have to do to set it to ONLY run on the GPU? The windows client has the "advanced mode" window where I can remove the CPU.

Link to comment
  • 10 months later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.