Because my computer is broke at the moment, I obviously haven't been folding. Once my computer is up-and-running again, installing folding will be one of my priorities, and that will help boost the team's points again.

I had some bad news in work though; the tech people found folding@home on SOME of the computers that I installed it on and deleted the program.

There's at least one computer that it is still running on though.

...

I've got no problems on my end with the extremeoverclockers site. Like Stumpy i've been keping track of our points. I've also noticed that (As usual) we're returning a lot of units that are pointless - reading what you just said Chris, it sounds like it is you that is returning all those broken units!

Are you SURE that folding@home is working correctly for you? I've only ever had one legitimate unit fail in all the months and years that i've been folding - but *someone* is returning about 20-40 incomplete units a day!

That's a lot of wasted processing and time, especially since we're not getting any points for it...

Probably is but I've no ideas why, ok I know that OpenGL games don't work with it, neither does P2P file sharing or some other programs but other times it'll just fail for no reason.

They didn't realize it was you did they MOE?

Yeah...but I denied all knowledge.

Sorry I haven't been doing a lot of folding these days guys. My computer was running really slow, so I recently did a reformat. I then decided I'd put Windows Vista as my primary OS this time. The install went fine, but Punkbuster in Battlefield 2142 didn't like it so I was always kicked out of the games. So I did another reformat of the OS partition and put XP MCE back on and everything is working again. I'm at work right now, but I'll reinstall F@H when I get home and my GFX card can get back to work!

Is it possable that when I was having problems with mine that it sent broken Wu's in? Dont see how though as I could not send anything.. Anyhow I'm back and running normal now. too bad to cost me 6 days.
I'm now at 12,600 of 20,000... It would be nice if they would not give us all the lowest point WU's they have. but folders can't be choosy I guess.

Wow yesterday some one managed to complete 636 units yesterday totaling a grand total of 0 points! lol

Oh well at least you have folding installed and trying to fold!

How can we find who is doing that?

here is my log file. does anyone have one that looks like this?

Launch directory: C:\Program Files\Folding@Home


[00:06:04] - Ask before connecting: No
[00:06:04] - User name: Winterhawk7 (Team 46574)
[00:06:04] - User ID: 348B0B6610F88B59
[00:06:04] - Machine ID: 1
[00:06:04]
[00:06:04] Loaded queue successfully.
[00:06:04] Initialization complete
[00:06:04] + Benchmarking ...
[00:06:08]
[00:06:08] + Processing work unit
[00:06:08] Core required: FahCore_78.exe
[00:06:08] Core found.



[00:06:08] Working on Unit 00 [January 22 00:06:08] <--<--<--<--------\/



[00:06:08] + Working ...
[00:06:08]
[00:06:08] *------------------------------*
[00:06:08] Folding@Home Gromacs Core
[00:06:08] Version 1.90 (March 8, 2006)
[00:06:08]
[00:06:08] Preparing to commence simulation
[00:06:08] - Looking at optimizations...
[00:06:08] - Files status OK
[00:06:08] - Expanded 79014 -> 420497 (decompressed 532.1 percent)
[00:06:08]
[00:06:08] Project: 2125 (Run 270, Clone 2, Gen 19)
[00:06:08]
[00:06:10] Assembly optimizations on if available.
[00:06:10] Entering M.D.
[00:06:32] (Starting from checkpoint)
[00:06:32] Protein: p2125_lambda_5way_melt_4_10011


Please let me know. If I am the only one with this I'm going to shut down and pull FAH out and reload it.. Also it would be a good idea if someone brought this to the attention of the Standford people. They might be able to trace out who is not getting anywhere.

Actually I am setting my folding to Pause for the day. Since you al can see who is doing untis or not. this is a good way to tell.

I have paused my work at 9:52 a.m. (U.S. mountain time) GMT -7 .....

I'll check back here tomorrow and see if the problem is still happening. If so then I will know it is not me.

Nope never seen anything like that on mine - but my log file is pretty weird anyway!

Btw, thanks to those who mentioned possible solutions but EOC is working fine now and I have no idea how! Just does lol

there are alot of wu's turned in at http://folding.extremeoverclocking.com/ ... s=&t=46574

1411 work units?! Ok, something is REALLY going wrong for someone somewhere...

Trekboy, I can't see anything wrong with Winterhawk's Log file - mine looked just like that when I was running folding on my computer, so could you post a copy of your log file for us?

I'm not trying to point a finger or anything, but your comment that your log file doesn't look like Winterhawk's surprises and confuses me.

Of course, if you're running the graphics card client and none of the rest of us are, you're excused.

just looked on the forums for a reason why someone is returning units worth nothing but instead found this. It explains the points difference between the 20,000 units and the rest.


"p1497 and other like it are bigWUs, which means they have a 100% bonus applied to them. This means the benchmark machine would actually score 220PPD on them. These WUs also benefit greatly from a large low latency cache. Anything larger than the 512Kb L2 cache that the benchmark machine has will result in a dramatic speed-up in processing.

p2125 by comparison, is a "bog standard" gromacs WU, with no bonus applied, and is limited by raw speed and not by cache, or memory bandwidth. It runs at more-or-less 110PPD on the benchmark machine.

So, 2125 is a standard WU, which is getting "normal" points/speed, whereas 1497 gets a 100% bonus in points, and processes much faster on large cache CPUs. "

"p1497 is worth 364 points. It has a bonus applied to it because it is a bigWU, which means that by the benchmark it should be worth only 182 points. It processes really fast on your hardware because you have much more cache than the benchmark machine.

If you ran p1497 on the benchmark machine with the points as they are (364 and 392), the 1497 would get 220PPD and the 2125 would get 110PPD, since the 1497 has the 100% bonus applied to it.
On your machine you are getting even higher PPDs because the 1497 likes large caches and runs even faster. "

heres the link to the forum page if you want it....

http://forum.folding-community.org/fpost162399.html#162399

Tall,

Thats neat, but how does it help to find what is causing all the 0 credits that the team is getting? I mean to say , is it me causing this?

Yeah I know, I got side tracked. But I've just posted something on the forums with a picture of our stats page. Wasn't too sure how to explain it!

Hopefully some one should reply soon. Theres alot of messages on the forums and most are too technical for me to understand.

Probally get some rude comments about no searching......blah blah blah

Edit: Oh just noticed we're not the only ones with this poblem.Some other guys asking if they can delete these pointless units.

Edit 2: just got a reply and here's what i got.

"Someone has an unstable CPU or GPU client that is burning through work units, generating a lot of Early Unit End errors (EUEs) likely because the CPU or GPU is overclocked too much, or not setup correctly. This is hurtful to the project, and should be corrected as soon a possible! "

Trekboy, you mentioned once apon a time that you think it was you coz of the p2p? is there any chance your overclock coz I know you have amazing knowledge of pc's!

.....

Edit 3: Oh sorry winter i never noticed your question. No I dont think its you causing the problem. Obviously i cant be sure but we have been having this problem for a while now (before you started to fold for us) and its only started to get really bad in that last couple of days.

i don't think it's me here's everything in my log file.

# Windows Graphical Edition ###################################################
###############################################################################

Folding@Home Client Version 5.03

http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: C:\Program Files\Folding@Home


[18:07:38] - Ask before connecting: No
[18:07:38] - User name: qaz79 (Team 46574)
[18:07:38] - User ID: 18AEDF7C71B3B07F
[18:07:38] - Machine ID: 1
[18:07:38]
[18:07:38] Loaded queue successfully.
[18:07:38] Initialization complete
[18:07:38] + Benchmarking ...
[18:07:41]
[18:07:41] + Processing work unit
[18:07:41] Core required: FahCore_78.exe
[18:07:41] Core found.
[18:07:41] Working on Unit 03 [January 22 18:07:41]
[18:07:41] + Working ...
[18:07:43]
[18:07:43] *------------------------------*
[18:07:43] Folding@Home Gromacs Core
[18:07:43] Version 1.90 (March 8, 2006)
[18:07:43]
[18:07:43] Preparing to commence simulation
[18:07:43] - Looking at optimizations...
[18:07:43] - Files status OK
[18:07:43] - Expanded 79112 -> 420497 (decompressed 531.5 percent)
[18:07:43]
[18:07:43] Project: 2125 (Run 85, Clone 36, Gen 5)
[18:07:43]
[18:07:43] Assembly optimizations on if available.
[18:07:43] Entering M.D.
[18:08:07] (Starting from checkpoint)
[18:08:12] Protein: p2125_lambda_5way_melt_4_10011
[18:08:12]
[18:08:12] Writing local files
[18:08:13] Using table 1
[18:08:14] Completed 6520565 out of 20000000 steps (33)
[18:08:15] Extra SSE boost OK.
[19:04:27] Writing local files
[19:04:28] Completed 6600000 out of 20000000 steps (33)

Folding@Home Client Shutdown.


--- Opening Log file [January 22 21:49:04]


# Windows Graphical Edition ###################################################
###############################################################################

Folding@Home Client Version 5.03

http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: C:\Program Files\Folding@Home


[21:49:04] - Ask before connecting: No
[21:49:04] - User name: qaz79 (Team 46574)
[21:49:04] - User ID: 18AEDF7C71B3B07F
[21:49:04] - Machine ID: 1
[21:49:04]
[21:49:05] Loaded queue successfully.
[21:49:05] Initialization complete
[21:49:05] + Benchmarking ...
[21:49:08]
[21:49:08] + Processing work unit
[21:49:08] Core required: FahCore_78.exe
[21:49:08] Core found.
[21:49:08] Working on Unit 03 [January 22 21:49:08]
[21:49:08] + Working ...
[21:49:10]
[21:49:10] *------------------------------*
[21:49:10] Folding@Home Gromacs Core
[21:49:10] Version 1.90 (March 8, 2006)
[21:49:10]
[21:49:10] Preparing to commence simulation
[21:49:10] - Looking at optimizations...
[21:49:10] - Files status OK
[21:49:10] - Expanded 79112 -> 420497 (decompressed 531.5 percent)
[21:49:10]
[21:49:10] Project: 2125 (Run 85, Clone 36, Gen 5)
[21:49:10]
[21:49:10] Assembly optimizations on if available.
[21:49:10] Entering M.D.
[21:50:15] (Starting from checkpoint)
[21:50:20] Protein: p2125_lambda_5way_melt_4_10011
[21:50:20]
[21:50:30] Writing local files
[21:50:30] Using table 1
[21:50:55] Completed 6650080 out of 20000000 steps (33)
[21:51:10] Extra SSE boost OK.

Folding@Home Client Shutdown.


--- Opening Log file [January 23 00:40:53]


# Windows Graphical Edition ###################################################
###############################################################################

Folding@Home Client Version 5.03

http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: C:\Program Files\Folding@Home


[00:40:53] - Ask before connecting: No
[00:40:53] - User name: qaz79 (Team 46574)
[00:40:53] - User ID: 18AEDF7C71B3B07F
[00:40:53] - Machine ID: 1
[00:40:53]
[00:40:53] Loaded queue successfully.
[00:40:53] Initialization complete
[00:40:53] + Benchmarking ...
[00:40:56]
[00:40:56] + Processing work unit
[00:40:56] Core required: FahCore_78.exe
[00:40:56] Core found.
[00:40:56] Working on Unit 03 [January 23 00:40:56]
[00:40:56] + Working ...
[00:40:57]
[00:40:57] *------------------------------*
[00:40:57] Folding@Home Gromacs Core
[00:40:57] Version 1.90 (March 8, 2006)
[00:40:57]
[00:40:57] Preparing to commence simulation
[00:40:57] - Looking at optimizations...
[00:40:57] - Files status OK
[00:40:57] - Expanded 79112 -> 420497 (decompressed 531.5 percent)
[00:40:58]
[00:40:58] Project: 2125 (Run 85, Clone 36, Gen 5)
[00:40:58]
[00:40:58] Assembly optimizations on if available.
[00:40:58] Entering M.D.
[00:41:19] (Starting from checkpoint)
[00:41:21] Protein: p2125_lambda_5way_melt_4_10011
[00:41:21]
[00:41:30] Writing local files
[00:41:35] Using table 1
[00:42:00] Completed 6651326 out of 20000000 steps (33)
[00:42:10] Extra SSE boost OK.

Qaz, no it's not you. Having been to those forums way more then I should I have learned a couple things. Your Logs are clean..


Now if we can just get that post shortened by 90% lol

Since this has gone on for longer then I have been folding.. it could be that someone WHO is NOT folding, might have it stuck "open" as in they dropped it but did not remove it from there system and it is still trying to run the programs..

sorry i didn't realize that the log was a couple of days long.

If that's true, Winter, we might have a problem. Most of the other guys are not just not folding, they're uncontactable. (Does that make sense? 8O)

The only way to stop it for the team is to make a new folding team - which would be pointless and wouldn't actually solve the folding problem.

Trekboy, if it is your overclocked CPU, please get it checked out. You could be holding back on a cure for cancer or something! 8O

right the guys from the board say that the person responsible needs to post there FAH log onto the board so that some on can take a look at it and see if they can solve the poblem.

http://forum.folding-community.org/viewtopic.php?p=162468#162468


They also had another brain wave. Look at our team points to see who's been submitting lots of units! lol So I checked and it is Trek boy AND whiteraven.

http://vspx27.stanford.edu/cgi-bin/main.py?qtype=teampage&teamnum=46574

Wow two people? Who would have thought that?

Good luck gettng it fixed, as the techino babble some of those folks use is way beyond understanding.

[EDIT] Ringolds Back! Welcome back Ringold. I started a while after you were gone. No one knew what happened to you.

Rumors were that you were working on a different team. lol Moe is having computer problems, and the rest well were trying!

(Guess I'll never reach that top position) Darn! lol lol lol

Yeah im having computer problems at the moment too.

I keep getting windows and system 32 error messages so my computer crashes alot lately so I can run my laptop over night.

Cant reformat either until i finish an assignment due in in Monday!

Yeah, I knew Ringold was back - I PM'd him and he's the latest member of our X-Wars alliance.

Don't assume that means he will be folding for the team though - he has lots of commitments and folds for many different programs like Africa@home, not just folding@home.

Obviously it would be great if he joined the team again, but don't pressure him guys.

Good luck Stumpy! I know how you feel! I've been retyping my sisters coursework for her since my computer broke!

Before I folded for our team I folded for [H]ardOCP, the #1 team. They're handily whooping the Aussies, so, they didn't need me I figured.

I like to spread out my projects, so I did some BOINC projects for a while.. wow, looks like almost a year!

Of my two computers, the older one, a dual Athlon MP 2400 system, I'm leaving running BOINC projects. This one I'm putting back on F@H.

Next month I'll be building a new system, and I'll have F@H on it too, at least for a while.

Tall_Stumpy_One, I can help with F@H problems some times, but that sounds like broad system instability, not just F@H's fault, so that's beyond my tech support skillz.

Regarding the log I saw Qaz & Winter post.. I see no error message?! just a core shutdown message. Did you guys get help for it already on the official forums?

If anybody ever wants to stress-test their CPU, here's the tool I use when I'm tweaking an overclock:

http://sp2004.fre3.com/beta/beta2.htm

It autodetects if you had a single or dual core CPU, and then lets you run the common overclockers list of different stability checking apps.

The most applicable one for our purpose here is 'StressCPU - Stress CPU with Gromacs core'.

Let 'er run for up to a day, and if it can be done with zero errors, then your hardware should be stable.

If not, god help you, 'cause it's probably time to jump in to the BIOS. And if it comes out clean but you still have F@H problems, then I'm at a loss. :\

Edit: Before running orthos, stop F@H of course. Otherwise they'll just compete for CPU cycles and it may not pick up an error.

Wow, sorry guys, I think the multiple WU's was me, as you can see by going to the team stats page when you right click on F@H, last time I checked I was at 3000+ WUs, but I haven't folded since then because of that. I think I need to install the CPU client before running the GPU client. There's no install for the GPU client but this is what happened when I ran it on it's own last time so I think it's the same thing this time. I'll report on what happens after I do some fiddling with it.

I havent had a chance to fiddle with the GPU client myself; X800XL for now, and a 8800GTS next month (which they havent figured out how to fold on yet).

If you manage to get your GPU folding working you'd have a shot at knocking me out of first. Currently, with a dual-core, can manage one console client and the GPU client (which eats up one core itself).

How would you run the CPU and GPU at the same time though? It says "Only one copy of this program can be running at a time. This copy will now quit." whenever I try to run them both.

I remember reading *somewhere* that you can change the settings so that folding@home is assigned to a single core on your computer - and there is built-in support for this that can theoretically handle up to 8 cores - so you could have 8 copies of folding running at the same time! 8O

...Of course, that would put a hell of a lot of strain on your RAM etc., and I don't know how you would go about setting it up...

Because I do not have the fancy system, the problem I had was traced to Mocrosoft and the new IE-7, So now that everything is adjusted, I fold a tad slower, but it is working again. So I'm happy! lol

It typically has to be done with the console version. At one point during the setup it'll ask you for the machine id...

From a fresh install, here is how I think it works:
1) Grab a standard FAHConsole504 or whatever, stick it in a directory where you want it to live.
2) Grab a GPU version, console version (or GUI if you're sure to set the Cpu console version as Machine ID 2). Stick it in a different folder than the cpu client.
3) Make a shortcut to the standard cpu client. Edit the shortcut and add "-local -configonly" flags to the end of it.
4) Run the shortcut. Enter data as appropriate. When it asks for advanced settings, say yes. For CPU usage, 100% at idle priority. Machine ID, 2
5) Repeat process with GUI install. Machine ID 1, 100% idle, if the choices are available. If not, it should default to Machine ID 1, which is why the cpu client I had you make #2.
6) Go registry diving. Start -> Run, regedit. Navigate to the services, which should be HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services.
7) Locate F@H@C:+....etc...
Right click and Modify the ImagePath binary, and add "-local -forceasm" flags for the CPU client's entry. For the GPU client, "-local" should be fine.
9) At this point, maybe your various clients are running, maybe not. Go to the control panel, admin tools or whatever, and look in Services. Start them if they arent going.

That should be it. Check the log files for errors after a little while to make sure it's all humming along.

The most hard-core folders try to get two CPU clients plus the GPU client running in a tricky balance of doom -- idle/idle/low priority, pausing the GPU client to play video files, etc.

Long discussion here: http://www.hardforum.com/showthread.php?t=1105415

But they're not 100% certain themselves its worth the effort for a 3-client bonanza (on a dual core system).

The -local flag and seperate machine ID's are the critical part to ensure that they dont bang their heads in to each other and shut down as you saw.

Edit: I came across this tool made by a regular at [H]ardOCP's forums. It might make installing a breeze for dual-core users, or for anyone else that wants a cpu console service install done quick and easy without the journey through the DOS box. Buyer Beware, Your Mileage May Vary, if it turns small children in to zombies don't blame me: http://userpages.umbc.edu/~willm1/setup-fold/

ok, so if I wanted to just be able to run them manually, not as services, I would just add the "-local -forceasm" flags for the CPU and the "-local" flag for the GPU to the respective shortcuts right? (and take away the -configonly of course)

EDIT: Ok, got both clients up and running together, but the GPU client is still just sending EOEs. I'm going to try something and I'll post the results in a minute.

EDIT2: Well I found out the problem with the GPU client is the 7.1 Catalyst driver I had, so I rolled back to 6.10 (by rolled back I mean "installed over" ) Now the client has stopped submitting EOEs, but is still starting the "GUI Server". I'll let it run overnight and see how it is in the morning (and by morning I mean noon )

EDIT3: Good afternoon So it seems that the GPU Client is running ok now. It completed a WU overnight, I think. I'm not really familiar with the Console version but it went up to "Completed 100" and sent the results so I would assume that would mean it finished a WU. Not sure the size of it, but it took almost exactly 13 hours to complete. So now everything is up and running again on my end. Thanks for your assistance with setting up the CPU and GPU clients Ringold. I'm BACK!!!!

I think in the future it'll be slightly easier.. there's too many things that can go wrong at the moment with the GPU client being in beta -- like picky ass drivers. That's why the 8800GTS/GTX's cant fold yet -- no drivers at all!

Good to see its working. The only drawback running them the way you are is I think, if I'm not mistaken, the DOS windows stay up all the time.

If the reason is that its not something you want running except certain times when you'll allow it to do so for not making it a service there's a handy little trick for managing that such that when they're running they're in the background, but easily startable and stoppable.

For anyone elses near-future reference..

1) Let the client install itself as a service, and otherwise follow previous instructions for changing the registry entry to look like "-svcstart -local -forceasm" at the end of the image page
2) Control Panel -> Admin Tools -> Services -> F@H -> Change Startup Type to Manaul. Note the precise name of the service.
3) Go to desktop. Create 2 new txt file, such as 'FAH Start.txt' and 'FAH Stop.txt'
4) Insert following code in to FAH Start:

Where sc start tells Windows to start the service and the rest is the name of your F@H service.
5) Do the same for FAH Stop, except, sc stop
6) Change them from .txt to .bat.
7) Double click one to start the service, double click the other to stop it!

Note 1: I keep saying to add -forceasm but I should note that it's only good for those of us with processors with SSE/3dNow and, for some cores, SSE2. Hopefully that includes all of us, but an original Celeron may not play nice.
Note 2: If 'sc start' doesn't work, I used to use net start and net stop but a quick Google suggests netstart is old school. sc start is 1 character shorter, so if it works, um, saved your finger from depressing one less key.

I used that setup a couple years ago when I had aircooling and didn't want the CPU fans spooling up until I was ready to step away.

Hopefully Stanford gets on the ball and makes this whole darn process easier from top to bottom. Seems rather old school and unnecessary for us enthusiasts to have to poke around in the bloody registry like it's 1999!

Anyway, congrats! You'll be outpacing me until I build a new comp in Feb or March no matter what I do, so, good luck.