Folding Forum

FAHControl PPD computation will be (or already are) broken : there has been a change in FAH API that is discontinuing the use of k-factor. All projects will use the same value which will be hardcoded and no longer published in project summary API.

The current symptom is that FAHControl will see a 0 for k-factor and will disable QRB for these projects, showing PPD based only on base point.

We may have to learn how to think using only base PPD for a while, but the issue should be fixed in a future update of FAHClient/FAHControl.

The credit estimates from work server after a WU upload and FAH stats are not affected by the issue and should display values with QRB.

I think that the problem is worse then that. I just go awarded 29000 point (base only) for Project: 14180 (Run 6, Clone 454, Gen 12). A clean run at normal speed without interruptions or errors
05:11:18:WU00:FS00:Sending unit results: id:00 state:SEND error:NO_ERROR project:14180 run:6 clone:454 gen:12 core:0x21 unit:0x000000110002894c5d3b5536557aab20.
But it ended so
05:31:58:WARNING:WU00:FS00:Exception: Failed to send results to work server: Transfer failed
05:31:58:WU00:FS00:Trying to send results to collection server
05:31:58:WU00:FS00:Uploading 8.98MiB to 128.252.203.4
05:31:58:WU00:FS00:Connecting to 128.252.203.4:8080
05:32:02:WU00:FS00:Upload complete
05:32:02:WU00:FS00:Server responded WORK_ACK (400)
05:32:02:WU00:FS00:Final credit estimate, 29000.00 points

Ubuntu. Sorry to high-jack but could it be a developing problem?

I have another 14180 finishing soon, client shows only base points for that as well. Folding set to finish now. I will update.

You're seeing a different problem, but it might have the same cause (k-factor being phased out) : it seems that when a WU is uploaded to a collection server instead of the work server that issued it, it doesn't apply the QRB ...

I guess the communication of projects settings between work server and collection server is broken ...

Is there a reason for discontinuation of k-factor? I realise I wasn't plugged in into fah community for a long time. Is there some other system in the works?

I guess that since NaCl has stopped (which didn't use QRB) and because all current (and probably future) projects use 0.75 for k-factor, there's no need to maintain this as a variable parameter ...

No other system currently in the works. Essentially the k-factor is not going away, but becoming a constant in the QRB equation. All projects done the last several years have used the same value, 0.75, and it has been determined that none will use a different value in the future.

So I take it very low PPD in FahControl is just Fahcontrol not adjusting to new setting?

Yes, the servers know the correct value so so stats won't be affected.

There is still a little communication issue between work servers with new software version and collection servers with an older version, but it should clear when all server will run the latest version.

Nice, thanks for the info

And will they address the missing points for already completed projects?

Yes, a manual recredit will probably occur ... but we don't know when someone will have some free time to do so, since it's not a high priority task ...

HFM received an update to fix the displayed PPD : viewtopic.php?f=14&t=9903&start=630#p308260

Goodness, I was really concerned when my PPD dropped. What a time to switch usernames and riser cabling. Any idea how this affects cure coin?

The actual points that are added to the database had some problems for a while but that appears to have been corrected going forward even if they appear to be reduced through certain viewing avenues. The expected re-credit, whenever it happens, will add a lot of points for everybody in a very short period of time.

tmontney wrote:Any idea how this affects cure coin?

[I am just a user here, not any kind of authority]

They do not support cure coin here, and do their best to never mention it. Do not expect any official reply.

Not sure if this is a coincidence but I had a funky crash the other day that required me to restore part of my C drive. When I fired F@H up I had every GPU slot cycle through failed downloads in repeat rapid succession until the slot failure message showed. This was only on my GPU slots, the CPU slot continued to work properly.

Deleting the work folder had no effect. Not knowing what else to do, I deleted the GPU slots and created new ones. No other changes were made and I kept the same GPU slot options as before (client-type=advanced, max-packet-size=big).

This got everything back up and running stable, except that FAHControl is reporting Estimated Credit the same as Base Credit, normally the former would be larger. As a result the Estimated PPD and Points Per Day are also WAY OFF. Checking my actual points however I can tell that the results are being reported and credited correctly. I have tried restarting FAHControl and the system but the issue persists.

FAHControl is the tool I use to check and make sure my folding rig is set up and working properly. I am hoping that the fix for this unfortunate glitch is addressed soon.