My CPU processes almost zero WU anymore

[BillAnderson]

For the las 4 days or so my GPU (GTX960) has processed the majority of WU usually 38,000 to 68,000 points for each work unit.

The CPU (Ryzen 5 2600x with 32GB 3000 RAM and between 6 and 10 threads allocated to F@H) processes almost no WU. In the past 36 hours only one WU has processed for 6,532 points

The log is filled with primarily 2 IPs for my CPU slot
WARNING:WU01:FS02:Failed to get assignment from '65.254.110.245:8080'
WARNING:WU01:FS02:Failed to get assignment from '18.218.241.186:80':

The odd entry will be for a different IP
WU01:FS02:Connecting to 40.114.52.201:8080
15:09:41:WARNING:WU01:FS02:WorkServer connection failed on port 8080 trying 80

Once I saw
WU01:FS02:Connecting to 155.247.164.214:8080
15:34:08:ERROR:WU01:FS02:Exception: 10001: Server responded: HTTP_SERVICE_UNAVAILABLE

The one success I have in CPU processing was
16:21:28:WU01:FS02:Connecting to 40.114.52.201:80
16:21:42:WU01:FS02:Downloading 2.14MiB

Have I done something to screw up the CPU processing? The only change I have made that I can recall is to install MSI Afterburner to set up a profile to control the GPU fan so it now lowers the GPU heat from 65C to 49C

I have tried setting CPU slots variously to 6, 8, 9, mostly I leave it at -1
Stumped

[Jesse_V]

Hey Bill,

I suspect that everything is just fine on your end. Those error messages, particularly the "failed to get assignment" and HTTP_SERVICE_UNAVAILABLE, are all caused by huge demand on the servers at the moment due to the massive rise in F@h's userbase over the past several weeks. I'd recommend setting the CPU slot exactly as you prefer and it should pick up workunits in time. The research teams are getting more projects in the queue and more servers online to keep up with demand, so hopefully that all settles out soon.

Related threads:
viewtopic.php?f=61&t=32424
viewtopic.php?f=61&t=33193

[jonault]

It doesn't seem likely that there's a configuration problem, but you can post your log file if you'd like folks to take a look for problems. If the client can access the servers to get work units for your GPU, it can do it for the CPU as well - you don't have separate network settings for each folding slot (unless the CPU & GPU in question are on two different computers). But CPU & GPU folding are two different things, and individual projects are only set up for one or the other. There have been times over the last week where all my CPU slots have been idle while the GPU slots have been active (well some of them anyway).

[BillAnderson]

Thanks jonault and Jesse_V for you replies. I'll just keep doin' what I'm doin'