Combined COVID-19 thread for DC updates

Gilthanis

[H]ard|DCer of the Year - 2014
Joined
Jan 29, 2006
Messages
8,718
Since we are getting multiple posts on FAH and Rosetta@home supporting COVID-19, I felt it would be nice for us to keep a singular point for updates on the topic so that people aren't trying to track down where things are posted. At the time of creating this thread, only two projects are working on it.

Rosetta@home (CPU only) - https://boinc.bakerlab.org/rosetta/forum_thread.php?id=13533#91785
Folding@home (CPU & GPU) - https://foldingforum.org/viewtopic.php?f=17&t=32124&sid=63bfec96e1b845c740487fdfeca113cd
TN-Grid (CPU only) - https://gene.disi.unitn.it/test/forum_thread.php?id=265
Still no update - GPUGrid - https://www.gpugrid.net/forum_thread.php?id=5078&nowrap=true#53922
BOINC@TACC (CPU only) - https://boinc.tacc.utexas.edu/forum_thread.php?id=44 -
Hello,

So far, we have found VirtualBox 6.0.14, 6.0.18, 6.1.2, 6.1.4 to work on Windows 10 clients. However, we have seen problems with 6.1.2, 6.1.4 on Ubuntu systems.

Sincerely,
the BOINC@TACC development team
WCG (CPU only for now but GPU work is now being BETA tested for 64 bit hosts for AMD/nVidia/Intel GPU's, Also runs on Android and Linux ARM devices) - https://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,42231 and https://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,42191
Completed the tests - Ibercivis (CPU only) - https://boinc.ibercivis.es/ibercivis/forum_thread.php?id=7#32
Dreamlab
sidock (CPU only) - sidock

Old posts on the topic
https://hardforum.com/threads/fah-has-covid-19-projects.1994003/
https://hardforum.com/threads/chinese-coronavirus.1992670/
https://hardforum.com/threads/could-folding-dc-help-find-a-cure-for-the-coronavirus.1992351/

Our Rosetta team - https://boinc.bakerlab.org/rosetta/team_display.php?teamid=5560
FAH team is Team 33
Our GPUGrid.net team - https://www.gpugrid.net/team_display.php?teamid=799
Our TN-Grid team - https://gene.disi.unitn.it/test/team_display.php?teamid=8
Our BOINC@TACC team - http://boinc.tacc.utexas.edu/team_display.php?teamid=10
Our WCG team - https://www.worldcommunitygrid.org/team/viewTeamInfo.do?teamId=BP5XNJBR9N1 but new accounts sign up with this link to autojoin our team - https://join.worldcommunitygrid.org?recruiterId=338542&teamId=BP5XNJBR9N1
Our ibercivis team - https://boinc.ibercivis.es/ibercivis/team_display.php?teamid=51
Our sidock team - https://www.sidock.si/sidock/team_display.php?teamid=7

If you want to set BOINC to pause work to allow for FAH to run when it has available work, please create a cc_config.xml file using Notepad and place it in the BOINC data directory. In Windows, it is by default located at c:\Program Data\BOINC\

Stop CPU & GPU work
<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
<report_results_immediately>1</report_results_immediately>
<dont_suspend_nci>1</dont_suspend_nci>
<exclusive_gpu_app>FahCore_21.exe</exclusive_gpu_app>
<exclusive_gpu_app>FahCore_22.exe</exclusive_gpu_app>
<exclusive_app>FahCore_a7</exclusive_app>
</options>
</cc_config>

Stop only GPU work
<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
<report_results_immediately>1</report_results_immediately>
<dont_suspend_nci>1</dont_suspend_nci>
<exclusive_gpu_app>FahCore_21.exe</exclusive_gpu_app>
<exclusive_gpu_app>FahCore_22.exe</exclusive_gpu_app>
</options>
</cc_config>
 
Last edited:
You can find some interesting information in this Reddit post -
 
I've never done this before but I'm in now and want to go all in. I noticed my GPU usage and CPU usage are only at around 40%. I have it set to the highest levels,even when I'm working, what gives? It's a 6600k and a 3GB 1060 on Win10.
 
I assume you are running folding at home, and have a slot for the CPU and GPU? What project is running in each slot? They have been having server troubles, so you may not be getting work.
 
I'm getting this in the logs

21:08:29:WU02:FS01:0x22:project: 11764 (Run 0, Clone 5587, Gen 6)
21:08:29:WU02:FS01:0x22:Unit: 0x0000000c80fccb0a5e71130adcad1454
21:08:29:WU02:FS01:0x22:Digital signatures verified
21:08:29:WU02:FS01:0x22:Folding@home GPU Core22 Folding@home Core
21:08:29:WU02:FS01:0x22:Version 0.0.2
21:08:29:WU02:FS01:0x22: Found a checkpoint file
21:08:41:WU02:FS01:0x22:Completed 850000 out of 1000000 steps (85%)
21:08:41:WU02:FS01:0x22:Temperature control disabled. Requirements: single Nvidia GPU, tmax must be < 110 and twait >= 900
21:08:41:ERROR:Receive error: 10053: An established connection was aborted by the software in your host machine.
21:09:01:WU00:FS00:Connecting to 65.254.110.245:8080
21:09:01:WARNING:WU00:FS00:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration
21:09:01:WU00:FS00:Connecting to 18.218.241.186:80
21:09:01:WARNING:WU00:FS00:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
21:09:01:ERROR:WU00:FS00:Exception: Could not get an assignment

Maybe it is server? I have it set to all causes. Pausing GPU doesn't matter, CPU just sits there at "ready" and usage is 1%. Turning GPU brings both up to 40%. Damn.
 
Yes, server. Unfortunately they are hammered (in a good way, tons of donors) right now and the servers are having trouble keeping up
 
Looks like you are doing what you need. The only other thing I could suggest is if you wanted to pretty much guarantee CPU work to run Rosetta@home using the BOINC software and let FAH have your GPU. (Rosetta only does CPU work) Then you could guarantee the CPU to ramp up as high as you let it. Just let us know if you want to run that too and if you need help with it. BOINC is a much bigger beast to fully learn and tweak to its fullest. However, Rosetta is pretty simple to set things up to run. Glad to have you on board.
 
I will add these lines in the first post as well for easy reference. However, if anyone wants to run BOINC work until their FAH client gets work to run, you can create a cc_config.xml file using Windows Notepad and add these lines to it:
<exclusive_app>FahCore_21.exe</exclusive_app>
<exclusive_app>FahCore_22.exe</exclusive_app>

So, for example, my cc_config.xml file would look like:

Stop CPU & GPU work
<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
<report_results_immediately>1</report_results_immediately>
<dont_suspend_nci>1</dont_suspend_nci>
<exclusive_gpu_app> >FahCore_21.exe</exclusive_gpu_app>
<exclusive_gpu_app>FahCore_22.exe</exclusive_gpu_app>
<exclusive_app>FahCore_a7</exclusive_app>
</options>
</cc_config>

Stop only GPU work
<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
<report_results_immediately>1</report_results_immediately>
<dont_suspend_nci>1</dont_suspend_nci>
<<exclusive_gpu_app> >FahCore_21.exe</exclusive_gpu_app>
<exclusive_gpu_app>FahCore_22.exe</exclusive_gpu_app>
</options>
</cc_config>

This config is telling BOINC to use all GPU's, report results immediately instead of waiting for the client to get around to it, not to suspend my NCI applications, and to pause when Windows sees the FahCore_21 or 22 executables in the task manager. This way it will run the FAH work units and then when you run out of them again, BOINC will resume as needed.

This file goes in the BOINC data directory. In Windows, the default location is c:\Program Data\BOINC and is a hidden folder. Other OS's can vary so if you are not sure where it is, please ask as someone would need to jump in to assist.

Adjusted it to only suspend GPU work units for BOINC. CPU should continue to run.
 
Last edited:
The folder in Linux is var\lib\boinc

and you have to open it as an administrator to get access to all the folders files and to be able to create documents inside it.
 
I thought the location could differ based on how you installed it on Linux. For example the repository installed it differently than if you downloaded the files from Berkeley and installed. Again, I'm not a Linux guru.
 
Thanks for the config, that is very helpful. FAH gets a few units a day but most of the time it is trying to connect to download or trying to upload.

Edit: I did have to use the command "exclusive_app" instead of "exclusive_gpu_app". Not sure why but it may be my system setup, I use an external GPU.
 
Last edited:
Yeah the servers are still getting hammered, despite the fact they they are now assigning 130k+ work units an hour
 
Still folding when I get a WU. R@H seems down right now so I just joined TN. Hope it helps :/
 
I do not fold with GPUs at the moment, but I installed BOINC inside a Docker container and BOINC is configured to pause using a cc_config.xml file similar to the one Gilthanis posted. My quick-n-dirty script is here: https://pastebin.com/CqTH1hq3. I'm currently using William Lam's OVA since it was so simple to deploy on existing ESXi hosts using ovftool. William posted sample scripts here: https://github.com/lamw/vmware-fah-automation
 
Another BOINC project jumping in the action.
Dear All:

Thanks for your continued support for the BOINC@TACC project!

Just wanted to give you all a heads-up that between today and next Monday, you would see an increase in the number of jobs submitted from BOINC@TACC. Majority of these jobs would be related to the SARS-COV2 related research.

Best Regards,
The BOINC@TACC Development Team

https://boinc.tacc.utexas.edu/forum_thread.php?id=44
 
Throwing in my rig for BOINC plus F@H with the recommended config file. Have BOINC at 70% CPU utilization. CPU at 73F max with stock cooler, will keep an eye on it. I think I need a new case with some sound dampening as the Corsair 200R is bloody loud. Work from home will be interesting with the noise, maybe just run at night/weekends.
 
That is pretty cool. Never visited that site before. But I doubt they would use that as an April fools joke. If it was, I certainly would never return.
 
WCG posted -
Hello,

I want to update everyone that we have started Alpha testing against the actual COVID-19 target. We have a sample set of small ligands to compare against COVID-19. This Alpha testing is on devices that are owned by members of the World Community Grid team. We are doing this to help give us an idea if the binaries work well and the work units do not crash. So far things are looking well as we work through issues. Alpha testing will hopefully be short, but one of the things we are testing is work unit sizing as different ligands take different times to complete, sometimes exponentially. We are getting a larger subset of ligands to test on Alpha later today to help with the sizing problem.

Thanks again for your interest and look forward to a BETA soon :p
-Uplinger
 
Last edited:
BOINC@TACC has had some work units going out. They seem to request 4 threads (locking them down) but only using 2. Keep and eye on it. Also, remember that the project requires Virtualbox to be installed on your PC.
 
Back
Top