Parps

Associate
Joined
1 May 2006
Posts
810
Location
Bristol, UK
The following members are about to be stomped by yours truely within the next week:

Steven_Bremner
dieddun
Panchevre
Halk
MrSheen
Hunter
neil3k
Iamgud

Looks also like I'm going to be stomped by Dirk_Trutwin :(
 
Stomping corpses isn't that exciting and you'll do a lot of it down that end of the team so it's hardly worth a parp. You're moving so slowly you're never going to pass a living cell occupant.
Freakish_05 said:
Looks also like I'm going to be stomped by Dirk_Trutwin :(
And dotterybot & hockeyplaya10 before that.

What client are you running? You seem to be getting through a few WU that aren't giving you more than one or two points which means it's not running properly.

Me, I'm celebrating my birthday by dumping 4800 points which finally gets me up to cell 36. :D
 
Snapshot said:
What client are you running?
I'm running 5.04 on my little celeron POS which also acts as a web server/mail server/file server/downoad box so there isnt really that much processing power left over for folding. But I set it up to do my part for the research. After all, the machine is on 24/7, it might as well be doing something useful aye? :p
 
Only if it's actually completeing whole wu's.

If it's EUEing then thats no good for the science and indicates theres instability with your computer :)
 
Don't rely on the stats to tell you if a WU completed properly. They just record the number of WUs you have sent and returned, whether complete or not. When my GPU client went wonky my WU count on stats leapt up by a couple of hundred before I caught it.

Have a look at the folding logs on your computer. Easiest way to see what is really going on can be utilities like f@hmon. That can be run on your normal PC, reading files from your server, provided the folding folder is shared read rights over the network.
 
Have a look at the table at the bottom of this stats page and you'll see that most of your WU are generating 0 points and completing quite quickly. As a complete WU should generate 57 points as an absolute minimum and, on your rig, will take days to complete, I'm sure you'll see that there's a problem.
 
Right then, assuming that there is a problem (I have no idea, all I did was follow the instructions in the sticky);
A. How do I fix it?
B. Is there any point in me continuing to fold?
 
I'm not the best guy for telling you how to fix the problem, although checking your CPU temps and lowering your overclock are probably the first things to do. However I had a few "early unit ends" over the past few days on PCs that have not had this problem in almost a year of folding. I can only guess perhaps there were some duff WUs.

There is good reason to carry on folding. You will probably find this problem goes away by itself or you just need to lower your overclock slightly. The units you are getting to the end of are giving you some very good points and by folding you are helping a good cause and some interesting science.

Take a look at the files FAHlog.txt and FAHlog-Prev (if you have one) in your FAH folder, you should be able to get a feel for what the folding software is actually doing in the background while it folds. Look for runs of numbers where the total doesn't get to 100 before trying to send back to Stanford. This is a WU that has not been processed correctly and won't get you more than 1-2 points worth of credit.
 
Is this on the Celeron machine which you use as a fileserver etc?

It may just be getting a tad too warm (time to clean out the dusty fuzz perhaps?) or if there is an overclock it may be worth backing it off a bit.
It should be able to run Folding fine as long as the system can handle the extra heat from running 100%. I know it fell over the other day when you posted about the sigs - was it doing that before you started running F@H on it?
 
This is running on the Celeron.
I've not got any overclocks on it.
I've checked the logs and for the last 5 or so days most of my WU's have been Early End Units :(
If I reduce the CPU Usage Required variable to like 70 would that help?
 
Well you could try that but it's probably best to track down the problem rather than just trying not to trigger it.

Use something like MBM5 to check the temperatures and it may be worth running Memtest either in Windows or download an ISO and burn to a disc.
 
Just seen on your website that the celeron is a 667Mhz jobbie. I think you should just remove FAH from the machine. It won't help your score by any noticable amount and will probably miss deadlines meaning hard work is wasted.
 
Sorry if some things I said seemed a bit odd. I was getting you confused with someone else. I thought you had folding running on some other PCs as well as the little celeron box.

Have you considered running FAH on another (more powerful) system?
 
verbal said:
Sorry if some things I said seemed a bit odd. I was getting you confused with someone else. I thought you had folding running on some other PCs as well as the little celeron box.

Have you considered running FAH on another (more powerful) system?
Up until tuesday it was just the little celly running. Then when we suspected errors I got both cores on my lappy running. But I checked the logs this morning and units seem to have stopped 10% of the way through, not bothered reporting anything and started new ones :confused: So I've just shut everything down until I have the time to get a more sustainable solution up and running :( Sorry Team.
 
Back
Top Bottom