No WUs available for CPU, unless I restart FAH

[kieranrjc]

Apologies if this is a dupe post, I wasn't able to see anything pertaining to this but maybe I'm just blind.

I've seen plenty about the shortage due to overwhelming demand (of which I'm a new joiner) however I've noticed that my GPU goes on and off regularly, receiving WUs.

However my CPU will do a single WU, and then get stuck on acquisition.

Fully exiting out (i've tried pausing/removing the slot to no avail) and restarting the client will cause my CPU to find one immediately, hinting that this may not be relating to the shortage.

Here's the log for the CPU slot starting a WU after re-opening FAH;

Code: Select all

08:48:23:WU00:FS00:Starting
08:48:23:WU00:FS00:Running FahCore: "C:\Program Files (x86)\FAHClient/FAHCoreWrapper.exe" C:\Users\Winge\AppData\Roaming\FAHClient\cores/cores.foldingathome.org/v7/win/64bit/avx/Core_a7.fah/FahCore_a7.exe -dir 00 -suffix 01 -version 705 -lifeline 10216 -checkpoint 15 -np 15
08:48:23:WU00:FS00:Started FahCore on PID 9468
08:48:23:WU00:FS00:Core PID:5796
08:48:23:WU00:FS00:FahCore 0xa7 started
08:48:24:WU00:FS00:0xa7:*********************** Log Started 2020-03-30T08:48:23Z ***********************
08:48:24:WU00:FS00:0xa7:************************** Gromacs Folding@home Core ***************************
08:48:24:WU00:FS00:0xa7:       Type: 0xa7
08:48:24:WU00:FS00:0xa7:       Core: Gromacs
08:48:24:WU00:FS00:0xa7:       Args: -dir 00 -suffix 01 -version 705 -lifeline 9468 -checkpoint 15 -np
08:48:24:WU00:FS00:0xa7:             15
08:48:24:WU00:FS00:0xa7:************************************ CBang *************************************
08:48:24:WU00:FS00:0xa7:       Date: Oct 26 2019
08:48:24:WU00:FS00:0xa7:       Time: 01:38:25
08:48:24:WU00:FS00:0xa7:   Revision: c46a1a011a24143739ac7218c5a435f66777f62f
08:48:24:WU00:FS00:0xa7:     Branch: master
08:48:24:WU00:FS00:0xa7:   Compiler: Visual C++ 2008
08:48:24:WU00:FS00:0xa7:    Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
08:48:24:WU00:FS00:0xa7:   Platform: win32 10
08:48:24:WU00:FS00:0xa7:       Bits: 64
08:48:24:WU00:FS00:0xa7:       Mode: Release
08:48:24:WU00:FS00:0xa7:************************************ System ************************************
08:48:24:WU00:FS00:0xa7:        CPU: AMD Ryzen 7 3800X 8-Core Processor
08:48:24:WU00:FS00:0xa7:     CPU ID: AuthenticAMD Family 23 Model 113 Stepping 0
08:48:24:WU00:FS00:0xa7:       CPUs: 16
08:48:24:WU00:FS00:0xa7:     Memory: 31.93GiB
08:48:24:WU00:FS00:0xa7:Free Memory: 24.65GiB
08:48:24:WU00:FS00:0xa7:    Threads: WINDOWS_THREADS
08:48:24:WU00:FS00:0xa7: OS Version: 6.2
08:48:24:WU00:FS00:0xa7:Has Battery: false
08:48:24:WU00:FS00:0xa7: On Battery: false
08:48:24:WU00:FS00:0xa7: UTC Offset: 1
08:48:24:WU00:FS00:0xa7:        PID: 5796
08:48:24:WU00:FS00:0xa7:        CWD: C:\Users\Winge\AppData\Roaming\FAHClient\work
08:48:24:WU00:FS00:0xa7:******************************** Build - libFAH ********************************
08:48:24:WU00:FS00:0xa7:    Version: 0.0.18
08:48:24:WU00:FS00:0xa7:     Author: Joseph Coffland <[email protected]>
08:48:24:WU00:FS00:0xa7:  Copyright: 2019 foldingathome.org
08:48:24:WU00:FS00:0xa7:   Homepage: https://foldingathome.org/
08:48:24:WU00:FS00:0xa7:       Date: Oct 26 2019
08:48:24:WU00:FS00:0xa7:       Time: 01:52:30
08:48:24:WU00:FS00:0xa7:   Revision: c1e3513b1bc0c16013668f2173ee969e5995b38e
08:48:24:WU00:FS00:0xa7:     Branch: master
08:48:24:WU00:FS00:0xa7:   Compiler: Visual C++ 2008
08:48:24:WU00:FS00:0xa7:    Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
08:48:24:WU00:FS00:0xa7:   Platform: win32 10
08:48:24:WU00:FS00:0xa7:       Bits: 64
08:48:24:WU00:FS00:0xa7:       Mode: Release
08:48:24:WU00:FS00:0xa7:************************************ Build *************************************
08:48:24:WU00:FS00:0xa7:       SIMD: avx_256
08:48:24:WU00:FS00:0xa7:********************************************************************************
08:48:24:WU00:FS00:0xa7:Project: 13860 (Run 0, Clone 699, Gen 14)
08:48:24:WU00:FS00:0xa7:Unit: 0x000000120d5a98395e72df9e982dcbb8
08:48:24:WU00:FS00:0xa7:Reading tar file core.xml
08:48:24:WU00:FS00:0xa7:Reading tar file frame14.tpr
08:48:24:WU00:FS00:0xa7:Digital signatures verified
08:48:24:WU00:FS00:0xa7:Calling: mdrun -s frame14.tpr -o frame14.trr -x frame14.xtc -e frame14.edr -cpt 15 -nt 15

Here's the failure to acquire a new WU during the end of the previous WU (repeated forever afterwards);

Code: Select all

09:25:35:WU02:FS00:Connecting to 65.254.110.245:8080
09:25:35:WARNING:WU02:FS00:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration
09:25:35:WU02:FS00:Connecting to 18.218.241.186:80
09:25:36:WU02:FS00:Assigned to work server 168.245.198.125
09:25:36:WU02:FS00:Requesting new work unit for slot 00: RUNNING cpu:15 from 168.245.198.125
09:25:36:WU02:FS00:Connecting to 168.245.198.125:8080
09:25:36:ERROR:WU02:FS00:Exception: Server did not assign work unit
09:25:36:WU02:FS00:Connecting to 65.254.110.245:8080
09:25:37:WARNING:WU02:FS00:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration
09:25:37:WU02:FS00:Connecting to 18.218.241.186:80
09:25:37:WARNING:WU02:FS00:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
09:25:37:ERROR:WU02:FS00:Exception: Could not get an assignment

I am still using the default configuration from initally installing FAH, and maybe missed something along the way

Any help would be appreciated.

[Neil-B]

One thing I have spotted is you have 15cores on your cpu slot -which you will find a fair few threads suggest is avoided so you might want to use the advanced control to configure the slot down to 12 (14 and 13 are also no necessarily good).

If you need more info on how to find/use the advanced control reply and I'll try and describe.

Other than that (unless I have missed something) it looks like you are just waiting in the queue (with many others) for CPU WUs ... What you are doing by exiting is probably shortening the retry counter which gets longer the more tries it attempts (so as not to overload the already overloaded servers with too many requests ... You can always reset the countdown timer in advanced control by pausing the cpu slot (right click on cpu slot line and choose pause ... watch the logs for a series on lines about a new config then go back and right click on the cpu slot and select fold.

This means you won't be interrupting GPU folding ... but please don't try this too much (perhaps only once retry times are more than an hour as it can just make the overload worse.

The may be an issue but it is worth changing you CPU slot cores and then leaving it for a few days - if you don't find you are picking up CPU WUs (and there can be hours or even for some people days) then repost and someone have be able to troubleshoot further.

[kieranrjc]

Thanks for the response, I've dropped it down to 12 cores for the time being, just to see if it helps.

Out of interest; when it was automatically choosing it would choose 15, though folding would still pin all 16 cores. Is this number a zero-index, so I should set it to 11 cores for 12, for example?

I had considered the timer reset, but I'd already tried a semi-sanitary test: I waited for ~6 hours just doing gpu, and then i'd let the next attempt roll through to no avail, and then restart FAH immediately finding a CPU WU. This could purely be chance, but it did feel a bit too repeatable.

[davidcoton]

The relevant thread count is what will actually fold. Multiples of prime numbers over 5 give problems, multiples of 5 don't work on all projects. So you want to set a count that will work. 16 would be good, but one thread is reserved to feed the GPU, so maximum 15 are available for CPU folding. If that doesn't get work, the next number down that stands a chance is 12.

You could then set up a second CPU slot with 3 threads, if you want full utilization

[Neil-B]

A GPU can fairly consistently use one CPU - So I guess what you had been seeing was 16 cores pretty much flat out but with 15 on the CPU slot and 1 working to keep GPU happy.

12 may be a good number as I believe there are a number of projects making their way through the process that are limited to maximum 12 cores ... I saw a couple of these hit my server (running advance) - they stepped my 30 and 24 cores slots down to 12 (which was cool).

The team have posted that restarts are not necessary (and aren't in themselves a magic fix for the server load issues) and are no different to pause/fold but if that is what you wish to do.

[saad3000]

I am running a 6 vCPUs and not yet received any WUs, its frustrating.

ERROR:WU00:FS00:Exception: Server did not assign work unit

keeps retrying with no luck.

[Joe_H]

I am running a 6 vCPUs and not yet received any WUs, its frustrating.

ERROR:WU00:FS00:Exception: Server did not assign work unit

keeps retrying with no luck.

Please post the first 200 or so lines of you log so we can see what type of work your system is requesting.

[saad3000]

Same problem with 16 vCPU:(

Code: Select all

16:48:58:***********************************************************************
16:48:58:<config>
16:48:58:  <!-- Folding Slot Configuration -->
16:48:58:  <client-type v='bigadv'/>
16:48:58:  <gpu v='false'/>
16:48:58:
16:48:58:  <!-- HTTP Server -->
16:48:58:  <allow v='127.0.0.1 0.0.0.0/0'/>
16:48:58:
16:48:58:  <!-- Remote Command Server -->
16:48:58:  <password v='*******'/>
16:48:58:
16:48:58:  <!-- Slot Control -->
16:48:58:  <power v='full'/>
16:48:58:
16:48:58:  <!-- User Information -->
16:48:58:  <team v='TEAM NUMBER'/>
16:48:58:  <user v='FAH@HOME'/>
16:48:58:
16:48:58:  <!-- Web Server -->
16:48:58:  <web-allow v='0.0.0.0/0'/>
16:48:58:
16:48:58:  <!-- Folding Slots -->
16:48:58:  <slot id='0' type='CPU'/>
16:48:58:</config>
16:48:58:Switching to user fahclient
16:48:58:Trying to access database...
16:48:58:Successfully acquired database lock
16:48:58:Enabled folding slot 00: READY cpu:16
16:48:58:WU00:FS00:Connecting to 65.254.110.245:8080
^[[93m16:48:59:WARNING:WU00:FS00:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration^[[0m
16:48:59:WU00:FS00:Connecting to 18.218.241.186:80
16:49:00:WU00:FS00:Assigned to work server 155.247.164.214
16:49:00:WU00:FS00:Requesting new work unit for slot 00: READY cpu:16 from 155.247.164.214
16:49:00:WU00:FS00:Connecting to 155.247.164.214:8080
^[[91m16:49:01:ERROR:WU00:FS00:Exception: 10001: Server responded: HTTP_FORBIDDEN^[[0m

Mod Edit: Added Code Tags - PantherX

[saad3000]

solved it by disabling iptables! thanks

[saad3000]

11:18:09:WU00:FS00:Requesting new work unit for slot 00: READY cpu:16 from 150.136.14.110
11:18:09:WU00:FS00:Connecting to 150.136.14.110:8080
11:18:09:ERROR:WU00:FS00:Exception: 10001: Server responded: HTTP_FORBIDDEN
11:19:09:WU00:FS00:Connecting to 65.254.110.245:8080
11:19:09:WU00:FS00:Assigned to work server 168.245.198.125
11:19:09:WU00:FS00:Requesting new work unit for slot 00: READY cpu:16 from 168.245.198.125
11:19:09:WU00:FS00:Connecting to 168.245.198.125:8080

Getting alot of forbiddens with Client type v=bigadv

[Neil-B]

not sure about the HTTP_FORBIDDEN error (very inexpert eye might guess firewall?) … but as a passing aside I believe v=bigadv is no longer a client type nearest is v=advanced … I guess that could be the issue but doubt it (again just a guess)

[saad3000]

According to the configuration guide it is:
https://foldingathome.org/support/faq/i ... ion-guide/
Big Advanced

v6: -bigadv
V7: client-type bigadv

[Neil-B]

See last post of this thread for an "update" … viewtopic.php?f=106&t=32706&p=314532 … it includes link to https://foldingathome.org/2015/01/31/bi ... anuary-31/ ... it may not cause a fault in the client and may default across to advanced but as a thing it has been deprecated … it is possible that big-adv still does something but that is not what posts I have seen imply

[saad3000]

thanks!

[Neil-B]

fwiw … advanced has kept my 24core and 32core slots relatively busy