Go Back   Hardware Canucks > HARDWARE CANUCKS COMMUNITY > HardwareCanucks F@H Team

    
Reply
 
LinkBack Thread Tools Display Modes
  #1 (permalink)  
Old April 25, 2009, 09:03 AM
lemonlime's Avatar
Hardware Canucks Reviewer
F@H
 
Join Date: Jul 2008
Location: Greater Toronto Area
Posts: 530
Default GPU2 woes..

Hi All,

My GTX260 has been producing over 6K PPD pretty reliably. I've had it overclocked to 700/1500/1200, and it appears to be very stable. I'm running the latest 182.50 drivers on Vista x64.

Lately, my GPU2 client keeps crapping out upon completion of the 300 pointers. I have been greeted by a 'FAH Core stopped responding' window for the third time now. Has anyone else seen this problem? I set all of the clocks back to default, but it appears to happen *after* a WU completes. Here is a portion of my log before the failure occured:

Code:
[07:18:57] Completed 97%
[07:19:34] Completed 98%
[07:20:12] Completed 99%
[07:20:49] Completed 100%
[07:20:49] Successful run
[07:20:49] DynamicWrapper: Finished Work Unit: sleep=10000
[07:20:59] Reserved 75620 bytes for xtc file; Cosm status=0
[07:20:59] Allocated 75620 bytes for xtc file
[07:20:59] - Reading up to 75620 from "work/wudata_00.xtc": Read 75620
[07:20:59] Read 75620 bytes from xtc file; available packet space=786354844
[07:20:59] xtc file hash check passed.
[07:20:59] Reserved 15168 15168 786354844 bytes for arc file=<work/wudata_00.trr> Cosm status=0
[07:20:59] Allocated 15168 bytes for arc file
[07:20:59] - Reading up to 15168 from "work/wudata_00.trr": Read 15168
[07:20:59] Read 15168 bytes from arc file; available packet space=786339676
[07:20:59] trr file hash check passed.
[07:20:59] Allocated 560 bytes for edr file
[07:20:59] Read bedfile
[07:20:59] edr file hash check passed.
[07:20:59] Allocated 0 bytes for logfile
[07:21:00] Could not open/read logfile=<work/wudata_00.log>; Cosm status=-1
[07:21:00] GuardedRun: success in DynamicWrapper
[07:21:00] GuardedRun: done
[07:21:00] Run: GuardedRun completed.
[07:21:01] - Writing 91860 bytes of core data to disk...
[07:21:01] Done: 91348 -> 90109 (compressed to 98.6 percent)
[07:21:01]   ... Done.
[07:21:01] - Shutting down core 
[07:21:01] 
[07:21:01] Folding@home Core Shutdown: FINISHED_UNIT
[07:21:04] CoreStatus = 64 (100)
[07:21:04] Sending work to server
[07:21:04] Project: 5765 (Run 4, Clone 393, Gen 432)
[07:21:04] - Read packet limit of 540015616... Set to 524286976.


[07:21:04] + Attempting to send results [April 22 07:21:04 UTC]
[07:21:05] + Results successfully sent
[07:21:05] Thank you for your contribution to Folding@Home.
[07:21:05] + Number of Units Completed: 371

[07:21:09] - Preparing to get new work unit...
[07:21:09] + Attempting to get work packet
[07:21:09] - Connecting to assignment server
[07:21:10] - Successful: assigned to (171.64.65.106).
[07:21:10] + News From Folding@Home: GPU folding beta
[07:21:10] Loaded queue successfully.
[07:21:11] + Closed connections
[07:21:11] 
[07:21:11] + Processing work unit
[07:21:11] Core required: FahCore_11.exe
[07:21:11] Core found.
[07:21:11] Working on queue slot 01 [April 22 07:21:11 UTC]
[07:21:11] + Working ...
[07:21:12] 
[07:21:12] *------------------------------*
[07:21:12] Folding@Home GPU Core - Beta
[07:21:12] Version 1.19 (Mon Nov 3 09:34:13 PST 2008)
[07:21:12] 
[07:21:12] Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[07:21:12] Build host: amoeba
[07:21:12] Board Type: Nvidia
[07:21:12] Core      : 
[07:21:12] Preparing to commence simulation
[07:21:12] - Looking at optimizations...
[07:21:12] - Created dyn
[07:21:12] - Files status OK
[07:21:12] - Expanded 68533 -> 357580 (decompressed 521.7 percent)
[07:21:12] Called DecompressByteArray: compressed_data_size=68533 data_size=357580, decompressed_data_size=357580 diff=0
[07:21:12] - Digital signature verified
[07:21:12] 
[07:21:12] Project: 5762 (Run 2, Clone 262, Gen 16)
[07:21:12] 
[07:21:12] Assembly optimizations on if available.
[07:21:12] Entering M.D.
[07:21:18] Working on Protein
[07:21:19] Client config found, loading data.
[07:21:20] Run: exception thrown during GuardedRun
[07:21:20] Run: exception thrown in GuardedRun -- Gromacs cannot continue further.
[07:21:20] Going to send back what have done -- stepsTotalG=10000000
[07:21:20] Work fraction=0.0000 steps=10000000.
[07:21:20] Starting GUI Server
[07:21:24] logfile size=0 infoLength=0 edr=0 trr=23
[20:13:10] CoreStatus = FF (255)
[20:13:16] Client-core communications error: ERROR 0xff
[20:13:16] This is a sign of more serious problems, shutting down.
Reply With Quote
  #2 (permalink)  
Old April 25, 2009, 09:29 AM
Banned
F@H
 
Join Date: Jun 2007
Location: Edmonton
Posts: 1,628
Default

Yes, there seems to be an issue with cores unloading from cpu in Vista 64 as of late. Are you in Vista 64 as well?
Reply With Quote
  #3 (permalink)  
Old April 25, 2009, 09:42 AM
lemonlime's Avatar
Hardware Canucks Reviewer
F@H
 
Join Date: Jul 2008
Location: Greater Toronto Area
Posts: 530
Default

Quote:
Originally Posted by cadaveca View Post
Yes, there seems to be an issue with cores unloading from cpu in Vista 64 as of late. Are you in Vista 64 as well?
Yep. That would probably explain it. Guess I just need to be patient and wait for either a driver or core fix--and keep a close eye on the rig.
Reply With Quote
  #4 (permalink)  
Old April 25, 2009, 10:21 AM
Alwaysrun's Avatar
Allstar
F@H
 
Join Date: Sep 2008
Location: Qualicum Beach BC
Posts: 801

My System Specs

Default

sucks lemons. I haven't had any go south on me for awhile but I resolved my problems awhile back with the same situation as your currently witnessing by having the core priority set to lowest possible and not having "Do not lock cores to specific CPU" checked. Give that a shot or try setting the processor affinity diferently as some have had success. At your OC you should be getting 7500ppd like I do if your 260 is the 216sp version like I have. I get only around 7000ppd lately because I just can't keep away from playing Riddick: Dark Athena
Reply With Quote
  #5 (permalink)  
Old April 25, 2009, 10:45 AM
chrisk's Avatar
Folding Captain
 
Join Date: Jul 2008
Location: GTA, Ontario
Posts: 7,410

My System Specs

Default

Some people have had issues running F@H in Vista if they installed in the program files directory. You can try installing outside of the Program Files Directory (if you re-install, make sure to manually delete your work folders which reside in the Application data folder):
Quote:
Windows Vista GPU client:
C:\Users\(your user name)\AppData\Roaming\Folding@home-gpu
__________________
Fold for team #54196
Reply With Quote
  #6 (permalink)  
Old April 25, 2009, 10:53 AM
chrisk's Avatar
Folding Captain
 
Join Date: Jul 2008
Location: GTA, Ontario
Posts: 7,410

My System Specs

Default

Found this post in the foldingforums:
Folding Forum &bull; View topic - Checkpoint failure at start of new WU - CoreStatus FF

Try deleting the work files in the folder I mentioned above. Pande says that core error you got (Corestatus FF 255) means a file was not deleted properly. Delete your old works files and try again.
__________________
Fold for team #54196
Reply With Quote
  #7 (permalink)  
Old April 25, 2009, 12:20 PM
lemonlime's Avatar
Hardware Canucks Reviewer
F@H
 
Join Date: Jul 2008
Location: Greater Toronto Area
Posts: 530
Default

Quote:
Originally Posted by Alwaysrun View Post
sucks lemons. I haven't had any go south on me for awhile but I resolved my problems awhile back with the same situation as your currently witnessing by having the core priority set to lowest possible and not having "Do not lock cores to specific CPU" checked. Give that a shot or try setting the processor affinity diferently as some have had success. At your OC you should be getting 7500ppd like I do if your 260 is the 216sp version like I have. I get only around 7000ppd lately because I just can't keep away from playing Riddick: Dark Athena
Thanks for the tips, Alwaysrun. I'm running dual SMP vmware instances on the same box, so perhaps it is something to do with affinity. I may try to disable the affinity lock and see what happens. I suspect that my PPD is a bit lower than it should be, but my card is a 192sp model, so definitely won't be able to keep up with the 7500ppd like you get. So far, it finished a few WUs without an issue.

Quote:
Originally Posted by chriskwarren
..Some people have had issues running F@H in Vista if they installed in the program files directory..
This is interesting. I believe mine is installed in that location. I'll try to move it as well to see if it makes a difference

Quote:
Originally Posted by chriskwarren
Found this post in the foldingforums:
Folding Forum &bull; View topic - Checkpoint failure at start of new WU - CoreStatus FF

Try deleting the work files in the folder I mentioned above. Pande says that core error you got (Corestatus FF 255) means a file was not deleted properly. Delete your old works files and try again.
Thanks again, I'll give this a shot as well.
Reply With Quote
  #8 (permalink)  
Old May 5, 2009, 07:33 AM
lemonlime's Avatar
Hardware Canucks Reviewer
F@H
 
Join Date: Jul 2008
Location: Greater Toronto Area
Posts: 530
Default

Everything seemed to correct itself, but unfortunately, got another FF last night. Not sure how long it was sitting there after submitting a WU, but I'm going to have to keep a close eye on it. The only thing I didn't try yet is moving the folder out of 'Program Files' and to another location. I'm thinking this might be a Vista thing as my XP rig has no issues with this.
Reply With Quote
Reply


Thread Tools
Display Modes

Similar Threads
Thread Thread Starter Forum Replies Last Post
Best drivers to use for Gpu2 Dashock HardwareCanucks F@H Team 3 April 2, 2009 08:28 PM
GPU2 folding + 3d screensavers..... BrainEater HardwareCanucks F@H Team 22 March 14, 2009 09:05 PM
More nvidia woes? Jonwall Rumor Mill 36 February 15, 2009 09:07 AM
A Not So Small Issue With GPU2 Client SKYMTL HardwareCanucks F@H Team 14 September 7, 2008 10:32 AM
My Overclocking Woes... Cptn Vortex Overclocking, Tweaking and Benchmarking 20 February 8, 2008 08:40 PM