Anyone out there Folding@Home?

I couldn't get in to Priuschat earlier, so I posted a response in your original thread in the FAH community forum.

Since the Stanford STATS page is showing only 3 CPUs within the last 7 days, I think it's a problem with two of your clients, not just the 6.20 one.

Do you know how to look at your log? It's the FAH-log.txt file in your folding directory. You should be able to see what workunit your system is processing - a number, plus numbers for Gen, Clone and Run. You'd be able to see if it's processing the same workunit over and over again. My response in the FAH forum links to instructions on how to run the qfix utility and how to run the client with the 'send' parameter to force it to try to send anything that's stuck in your queue.

 

For goodness sakes when I read the title of this thread I thought it was a laundry folding service.

 

Downloading older client is available at Folding@home - DownloadOld

If you use a console client, you can try to add the following flag to your startup client. This page, Folding@home - WinConsoleInstall, show the flag option available for console client.

-verbosity 9 This option should be used if the user wishes to see more detailed client output

-send all send all results outstanding.

 

I ignored this thread for a long time when I first joined thinking it was one of those "work from home" stuffing envelopes jobs posts.

Just glad I finally read it to find out what it really is.

 

Thanks for the great replies guys. So I added the flags (it's actually -send x now instead of -send all). Here's what I get in my log:


# Windows CPU Console Edition #################################################
###############################################################################

Folding@Home Client Version 6.20

Folding@home - Main

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

Launch directory: C:\Folding software
Executable: C:\Folding software\[email protected]
Arguments: -verbosity 9 -send x

[15:25:47] - Ask before connecting: No
[15:25:47] - User name: prberg (Team 52533)
[15:25:47] - User ID: 4AF7BE5A16EBC4F7
[15:25:47] - Machine ID: 1
[15:25:47]
[15:25:48] Loaded queue successfully.
[15:25:48] Attempting to return result(s) to server...
[15:25:48] Project: 0 (Run 0, Clone 0, Gen 0)
[15:25:48] - Warning: Asked to send unfinished unit to server
[15:25:48] - Failed to send unit 00 to server
[15:25:48] ***** Got a SIGTERM signal (2)
[15:25:48] Killing all core threads

Folding@Home Client Shutdown.


Does that mean anything? Are my results stuck in my computer? Should I try Jeannie's suggestion (qfix)?

thanks,
-Peter

 

there, Ranchogirl! Come join our club, folding proteins at home for science - basic science intended to further cures and treatments for diseases! It's really very easy, we're a good group of people (all probably Prius owners on our Priuschat team), and we're working hard to break into the top 200 teams among the tens of thousands of teams participating in this project. Your contribution would be very welcome!

 

Actually, it's -send 1 to send what's in queue 1, -send 2 to send what's in queue 2, etc, or -send all to send everything that's in any of your queues.

Yes, your log says your computer is trying to fold Project 0, Run 0, Gen 0, Clone 0, which means it's not actually folding a valid work unit. ALL projects have a number, like 2665. Run can be 0, and I don't think you can have a Gen 0, but I think you can have a Clone 0.

So stop your client, go into your folding directory, remove the WORK folder and the queue.dat file, and start your client up again. This will make your client believe it's starting over entirely, like a 'first time' start, which is fine since this client hasn't actually done any work in the week or two it's been running, doing nothing.

Then start watching the log. The first thing it should do is connect to a server at Stanford and download a Workunit (aka Project) file. Then it may or may not need to download the 'core' - the particular piece of software that does the calculations. Then you'll see what Workunit, Run, Gen and Clone it downloaded. Then you'll see that you're on frame/step 0, 0% done.

If you've set up your client to do 'checkpoints' every 'few' minutes (default is 15), then every 'few' (15) minutes you'll see it saying it's writing the checkpoint file, until it's 1% done, and it writes another checkpoint file and tells you it's 1% done, and it continues on that way until you stop your client again, or until you get to 100% done.

Once you get to 100% done, it will try to connect to a server at Stanford to send your results, and then it will try to download a new Workunit/Project, and keep folding, this time using a different queue slot (so that if the 'send' doesn't work, it keeps the data in the 'old' queue slot instead of overwriting it - I think there are 8 queue slots.)

Check out your other clients, also, because the stats page shows you've had 4 CPUs running in the past 50 days but only 3 CPUs in the last 7 days. Assuming the one that was doing project 0 was your new client, 6.20, then one of your older clients is also stuck. DO try the qfix on that one and run your client with the -send all from a DOS window, as explained in the link in your original post in the FAH community forum. Since that client USED to work properly sometime more than a week ago and steadily for 6 weeks or so before that, you very well may have some valid data in your queue that should be 'fixed' and then sent. When you restart that client, depending on how it was hung up, it may be able to pick up where it last left off, or it may need to start with a new workunit/project (almost certainly if the last thing that happened was an Early Unit End, also called EUE).

Good luck with it, Peter - we definitely want your valuable contribution from all 5 of your systems, not just the 3!

 

Congrats, Peter - I see the FAH stats page showing all 5 of your cpu's again!

Chinalfr2 - I looked at some of the NVIDIA boards you suggested, and they all had a minimum requirement of a 400W power supply.

 

Jeannie, most OEM power supply output is under rated. Any OEM power supply can easily handle a GeForce 8800 GT, even if you have multiple hard drive and DVD-rom/writer on the OEM computer. For OEM, I mean HP, Dell, Compaq computer.

You do have to make sure your OEM Motherboard have a PCIE slot for the video card. If your power supply do not have the PCIE power connector, you can used the PCIE power adapter (enclosed with the Video card package).

My Dell Dimension 8400 power supply is rated 300W, but it actually can support up to 450W.

 

Except I still don't have a clue what it is about. Folding proteins? Huh? The posts sound like computer jargon. Waaaay over my head.

Sorry...

 

The science of it is way over my head, too - I'm not a molecular biologist. But Folding@Home is producing results in basic science, and those results are being used in other projects that are working directly on finding treatments and cures for diseases.

So, I contribute the time on my computer when the CPU wouldn't be doing anything - simple concept, and I'm hoping that my little bit helps in the overall goal of curing diseases.

 

Team Priuschat on Facebook. If you are folding with Team Priuschat and active on facebook.com, please install "F@H Protein Researcher" application and update your info.

On facebook world, Team Priuschat rank No. 23. In the world, Team Priuschat rank No. 197 (and climbing).

 

I plan to be contributing more soon. I'll be putting F@H on my boss' computer when she's next out of town (tomorrow). It gets very light usage because she's on the road so much, so it should be able to crank out some decent points.

 

Please ask for your boss permission first before you install the application. Don't get into trouble by installing the application without approval.

 

Thanks for kind thoughts. My post sounded "sneakier" than it was.

We're a small business that really functions like family. My boss is the mother of the guy who served as best man at my wedding. She just happened to be in her office yesterday, and her current opinion on computers is the less she knows/sees the better.

In addition to being one of the attorneys (and an employment attorney at that - so I concur heartily in your general advice) in our office, I also wear the Sys Admin hat to the extent anyone does, and have free reign over the computers. So, no worries.

Peace,

Dave

 

Check out my last few packets.
http://kakaostats.com/usum.php?u=1095859

The latest SMP, console, windows. Before I quit awhile back, smp & 24 hours of crunch would net a few thousand points. Now I got a 84 and a 119 points.
Sure hope it gets better.

 

it looks like you're getting single-core work units. Do you have -smp specified in the arguments for your console client, or did you include it in the 'advanced' options in your configuration?

My Windows Vista SMP console clients are getting project 2665 workunits worth 1920 points.

I'm running the console client with the DEINO option, since for me the MPICH option didn't give me 100% usage of the cores. Use the installation guide for instructions for the installation: Folding@home - WinSMPGuideDEINO


After you re-install the DEINO option, go to this site to update the latest .exe: Folding Forum • View topic - new 6.22beta2R3 public beta client (w/ ack fix) which is 6.22beta2R3. You'll pick up a version of the .exe with several fixes, including better handling of early end unit problems. Just replace the .exe you downloaded with the DEINO option.


Nice to see you folding again, Mark, and I look forward to your contributions!