Failed Stream for GPU Only

Good evening.

Long time folder, first time posting at your site.

Background: For the past several years I have been folding for Bjorn3d and am currently right around 47,000,000 points. I am getting ready to change teams and was hoping to join Team 32. Unfortunately, I have been having a problem that I can't seem to fix. I run three, older machines in my farm: A) i5-3470 w/ gtx 470, B) i7-950 (OC 3.9) w/ gtx 470, and C) q9550 (OC 3.6) w/ gtx 460. A & B run Win7 64bit and C is still chugging along on Vista 64bit. Generally, in early summer to about the first of November, I use the machines to render video so I shut down the clients. After the rendering time is over, I always do clean installs of the OS on the machines - just a habit, not sure why I do it. A problem arose after installing v7. None of the three machines can "read the stream". Copy of the log for one of them is below but it is similar on all three. What's odd is, I can fold on all three machines using the CPU only through the Chrome app. The results are not bad but I would like to get the GPU's up and running. I have an i7-2600k (OC 4.6) w/ gtx 670 ready to replace the q9550 but, until i can get this cleared up, there is no sense.

For the past two weeks, I've been reading all the postings I can find that discuss this issue but still can't quite get everything to work. I know any work is good work when it comes to folding but I would love to have everything running on all cylinders.

I could not get the log to show properly so I just copied a portion of it below. Sorry about that.

*********************** Log Started 2014-11-18T19:38:23Z ***********************
19:38:23:************************* Folding@home Client *************************
19:38:23: Website: http://folding.stanford.edu/
19:38:23: Copyright: (c) 2009-2014 Stanford University
19:38:23: Author: Joseph Coffland <[email protected]>
19:38:23: Args: --open-web-control
19:38:23: Config: <none>
19:38:23:******************************** Build ********************************
19:38:23: Version: 7.4.4
19:38:23: Date: Mar 4 2014
19:38:23: Time: 20:26:54
19:38:23: SVN Rev: 4130
19:38:23: Branch: fah/trunk/client
19:38:23: Compiler: Intel(R) C++ MSVC 1500 mode 1200
19:38:23: Options: /TP /nologo /EHa /Qdiag-disable:4297,4103,1786,279 /Ox -arch:SSE
19:38:23: /QaxSSE2,SSE3,SSSE3,SSE4.1,SSE4.2 /Qopenmp /Qrestrict /MT /Qmkl
19:38:23: Platform: win32 XP
19:38:23: Bits: 32
19:38:23: Mode: Release
19:38:23:******************************* System ********************************
19:38:23: CPU: Intel(R) Core(TM) i5-3470 CPU @ 3.20GHz
19:38:23: CPU ID: GenuineIntel Family 6 Model 58 Stepping 9
19:38:23: CPUs: 4
19:38:23: Memory: 3.90GiB
19:38:23: Free Memory: 1.68GiB
19:38:23: Threads: WINDOWS_THREADS
19:38:23: OS Version: 6.1
19:38:23: Has Battery: false
19:38:23: On Battery: false
19:38:23: UTC Offset: -5
19:38:23: PID: 4640
19:38:23: CWD: C:/Users/TRFB2/Documents
19:38:23: OS: Windows 7 Home Premium
19:38:23: OS Arch: AMD64
19:38:23: GPUs: 1
19:38:23: GPU 0: NVIDIA:2 GF100 [GeForce GTX 470]
19:38:23: CUDA: 2.0
19:38:23: CUDA Driver: 6050
19:38:23:Win32 Service: false
19:38:23:***********************************************************************
19:38:23:<config>
19:38:23: <!-- Folding Slots -->
19:38:23:</config>
19:38:23:Connecting to assign-GPU.stanford.edu:80
19:38:44:WARNING:Attempting to update GPUs.txt from assign-GPU.stanford.edu:80: Failed to read stream
19:38:44:Connecting to assign-GPU2.stanford.edu:80
19:39:06:WARNING:Attempting to update GPUs.txt from assign-GPU2.stanford.edu:80: Failed to read stream
19:39:06:Trying to access database...
19:39:06:Successfully acquired database lock
19:39:06:Enabled folding slot 00: PAUSED cpu:2 (not configured)
19:39:06:Enabled folding slot 01: PAUSED gpu:0:GF100 [GeForce GTX 470] (not configured)
19:39:10:16:127.0.0.1:New Web connection
19:39:15:24:127.0.0.1:New Web connection
19:39:22:Set client configured
19:39:23:WU00:FS00:Connecting to 171.67.108.200:8080
19:39:23:WU01:FS01:Connecting to 171.67.108.200:8080
19:39:23:WU00:FS00:Connecting to 171.67.108.200:8080
19:39:23:WU01:FS01:Connecting to 171.67.108.200:80
19:39:23:WU00:FS00:Assigned to work server 171.64.65.79
19:39:23:WU00:FS00:Requesting new work unit for slot 00: READY cpu:2 from 171.64.65.79
19:39:23:WU00:FS00:Connecting to 171.64.65.79:8080
19:39:23:WU01:FS01:Assigned to work server 140.163.4.231
19:39:23:WU01:FS01:Requesting new work unit for slot 01: READY gpu:0:GF100 [GeForce GTX 470] from 140.163.4.231
19:39:23:WU01:FS01:Connecting to 140.163.4.231:8080
19:39:24:WU01:FS01ownloading 4.83MiB
19:39:24:WU00:FS00ownloading 250.56KiB
19:39:25:WU01:FS01ownload complete
19:39:25:WU01:FS01:Received Unit: id:01 stateOWNLOAD error:NO_ERROR project:13000 run:972 clone:0 gen:64 core:0x17 unit:0x00000077538b3db75310ae1e24998dcd
19:39:25:WU01:FS01ownloading core from http://web.stanford.edu/~pande/Win32/AMD64/NVIDIA/Fermi/Core_17.fah
19:39:25:WU01:FS01:Connecting to web.stanford.edu:80
19:39:25:WU00:FS00ownload complete
19:39:25:WU00:FS00:Received Unit: id:00 stateOWNLOAD error:NO_ERROR project:10424 run:4083 clone:2 gen:2 core:0xa4 unit:0x000000030a3b1e734ccfa2ba7ab4ef1b
19:39:25:WU00:FS00ownloading core from http://web.stanford.edu/~pande/Win32/AMD64/Core_a4.fah
19:39:25:WU00:FS00:Connecting to web.stanford.edu:80
19:39:46:ERROR:WU01:FS01:Exception: Failed to read stream
19:39:46:ERROR:WU00:FS00:Exception: Failed to read stream

Thank you,

Paul

Welcome to OCF.!.
You say you reinstalled OSs; was your data logging kept separate & potentially needs deletion?

Everything was deleted. Now, Since the clean install, I have installed and un-installed the F@H software several times but have deleted the data with every new attempt.

What configuration flags do you set?

Actually nothing other than my passkey. I've changed up the installation several times but have never had the opportunity to put in anything other than the passkey. Prior to the last clean install, I was running separate CPU and GPU clients on each machine and was using flags on them. Unless I am missing something, I'm not seeing where I can load the flags anymore.

Two things to look at.

1) Is the date and time set correctly on the machine? Being off more than 5 minutes can sometimes cause an issue.

2) Some researching came up with the idea that the user account doing the folding may not have admin rights to all of the FAH directories.

Worth a shot.

Pablo 54.

I was getting the error message "Failed to read stream", for about a week. I tried a bunch of stuff.
This is my thread:
http://www.overclockers.com/forums/showthread.php/752201-Nvidia-GPUs-not-folding

The last couple of things I tr1ed were:
1) Remove the router from your internet connection. Connect one computer to the internet via your modem, see if that works.
2) If it does work, try resetting your modem to default settings (push the "reset" button while powered on for at least ten seconds). Then put the modem back into your network.

That seems to be what solved my problem.

Thanks for the tip. Checked the times, they were off by 2 minutes each. I switched all three to time-b.nist.gov. Is there a quick way to check for admin rights on my directories? I don't have any restrictions on anything but that does not mean I missed something.

Farwalker2u,

Funny you mention that. Your posting is what lead me to this site. You had done some stuff I had not tried before, so thank you. In between running the clients I did upgrade my equipment. I use TW internet and a cable card in an HD Prime box. I have my own modem and router and an HD antenna connected to two recorders. My gut feeling is there is something with the Silicondust software that might be interfering with my farm. Everything is connected via a GB switch, HD Prime is not. I did connect everything straight to the modem when I was trouble shooting but I never disconnected the HD Prime. So, last night I took one of my boxes to a friend's house, downloaded v7 and everything ran fine. Progress! Brought the box home, connected it directly to the modem, worked fine. Got up this morning, noticed none of my shows recorded in WMC - had "no signal" errors. That's never happened before. So, for the time being, it seems that if I take my Prime off line, I think I can fold. Going to double check that hypothesis tonight. Problem, no recorded shows. My wife may kill me. I'll post when I find out tonight if it works.

Thanks for the help.

Paul

Pablo 54 said:

Thanks for the tip. Checked the times, they were off by 2 minutes each. I switched all three to time-b.nist.gov. Is there a quick way to check for admin rights on my directories? I don't have any restrictions on anything but that does not mean I missed something.

Click to expand...

Nothing really special to check. I'm just referencing to see if you created a special "non" admin user to fold under. My guess is that this is not the issue.

Yes, as Farwalker mentioned, this is an internet related issue.

Connect your machine direct to your modem that you rent from your ISP or own and see if this works.
If so, it's a router or something between your modem and computer causing this.

Perhaps there is something going on with pande as of late to produce these errors.
Can't be coincide that you and Farwalker both have the same issue.

Sorry it took a bit. I was out of town. I agree with the internet related issue so I went for an easy shot first - didn't do anything but took my HD Prime off of my router. Everything works. Unfortunately, that does not help. I've had the same setup with the Prime for over a year. Why the glitch now? Before I even got to this point, I went to factory settings on the router. I hope I am not being paranoid but both of my kids are in college and use iMacs. Both have some cockeyed belief that their iMacs can't get viruses. Long story short, daughter comes home a month ago logs into our network, works the weekend, goes back to school then calls me a few days later to ask which anti-virus software I would recommend because she may have had a virus a few weeks ago. I won't tell you what I said but, anyway could I have something in my network causing this problem? I have not noticed any slowdowns or unexplained usage. I may drop cable in a month, price went up again.

Also, I take it the team number is 5297?

Pablo 54 said:

Also, I take it the team number is 5297?

Click to expand...

We are Team 32.