Win7 Update (KB973688) 11/26 Errors

Nagilum
Nagilum
Joined: 13 Feb 09
Posts: 12
Credit: 524868
RAC: 0
Topic 194640

Automatic Update Microsoft XML Core Services 4.0 Service Pack 2(KB973688)was loaded and severe client errors have occured ever since. Only 1 project listed on tasks now and stuck in waiting to run status. Restored from Boinc V6.10.18 to V6.10.17, but no change in current status. All other programs runing fine, but nothing as CPU intense as Boinc runs on this system. Error doesn't appear to be physical and smooth operation was reported until Windows Automatic Update occured. "Help!!"

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5893653
RAC: 1

Win7 Update (KB973688) 11/26 Errors

For the admins around here, it's an application error. He seems to be crashing in the CUDA FFT plan.

6.10.18

The system cannot find the path specified. (0x3) - exit code 3 (0x3)

Activated exception handling...
[17:47:52][5736][INFO ] Starting data processing...
[17:47:52][5736][INFO ] Using CUDA device #1 "GeForce 9800 GX2" (576.00 GFLOPS)
[17:47:52][5736][INFO ] Checkpoint file unavailable: status.cpt (No such file or directory).
------> Starting from scratch...
[17:47:52][5736][INFO ] Header contents:
------> Original WAPP file: p2030_54471_60901_0039_G47.57-00.32.S_2.wapp
------> Sample time in microseconds: 128
------> Observation time in seconds: 268.9792
------> Time stamp (MJD): 54471.704872685186
------> Number of samples/record: 512
------> Center freq in MHz: 1440
------> Channel band in MHz: 0.390625
------> Number of channels/record: 256
------> Nifs: 1
------> RA (J2000): 191921.025886
------> DEC (J2000): 124719.685045
------> Galactic l: 47.4705
------> Galactic b: -0.2582
------> Name: G47.57-00.32.S
------> Lagformat: 0
------> Sum: 1
------> Level: 3
------> AZ at start: 384.4993
------> ZA at start: 6.0539
------> AST at start: 0
------> LST at start: 0
------> Project ID: p2030
------> Observers: JD
------> File size (bytes): 16192559
------> Data size (bytes): 16179201
------> Number of samples: 2097152
------> Trial dispersion measure: 126 cm^-3 pc
------> Scale factor: 7398.64
[17:47:54][5736][INFO ] Seed for random number generator is -1152084332.
[17:47:54][5736][INFO ] Derived global search parameters:
------> f_A probability = 0.04
------> single bin prob(P_noise > P_thr) = 1.2977e-008
------> thr1 = 18.1601
------> thr2 = 21.263
------> thr4 = 26.2923
------> thr8 = 34.674
------> thr16 = 48.9881
[17:47:55][5736][ERROR] Error executing CUDA FFT plan (error code: 6)
[17:47:55][5736][ERROR] Demodulation failed (error: 3)!
17:47:55 (5736): called boinc_finish

]]>

Nagilum, have you tried a reboot of the machine?
Have you tried to update or reinstall your videocard & CUDA drivers?
Have you tried to run without the GPU on that system? 9To do so, edit your Einstein project preferences, change "Use NVIDIA GPU if present" from Yes to No and save changes. At the next contact with the project these preferences will be used and you will not load ABP1.

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5877
Credit: 118667223786
RAC: 19179148

RE: For the admins around

Message 95748 in response to message 95747

Quote:

For the admins around here, it's an application error. He seems to be crashing in the CUDA FFT plan.

....
[17:47:55][5736][ERROR] Error executing CUDA FFT plan (error code: 6)
[17:47:55][5736][ERROR] Demodulation failed (error: 3)!
17:47:55 (5736): called boinc_finish

....


Take a look at this message by one of the Devs, Oliver Bock. It seems it may be a problem with insufficient graphics memory at the time the problem occurred.

Cheers,
Gary.

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5893653
RAC: 1

RE: It seems it may be a

Message 95749 in response to message 95748

Quote:
It seems it may be a problem with insufficient graphics memory at the time the problem occurred.


Ah... that's useful.
Too bad the forum software here doesn't show people's videocards (with amount of memory and which drivers they run). Also too bad the error messages doesn't show the amount of memory still left over on the GPU. That's always very useful information to have.

Nagilum
Nagilum
Joined: 13 Feb 09
Posts: 12
Credit: 524868
RAC: 0

There is a possiblity the

Message 95750 in response to message 95747

There is a possiblity the CUDA errors were due to a new version of Boinc automatically starting-up at logon and me running other programs not realising Boinc was up. Previously I had released the msconfig from starting Boinc when my computer was first turned on because of interferance with audio editing programs while not realising Boinc was in BG. But I decided to get the full power of my quad Q6600 anyway and upgraded win7 32bit to the Win 7 64bit OS and started fresh. Once again I forgot to disable Boinc auto-start and noticed interferance while studio recording a friend today. But I've observed a more stable CUDA now that pause's correctly when I use my mouse across the screen and then resumes after 3 mins with no extra activity. With my primary apps setup correctly and asleep while Boinc is running, I'm going to resume crunching with the new 64bit system and see how things go. Afterall the system ran fine for a long period of time before I installed a version upgrade that replaced the start at logon command in msconfig. Thanks for your Help!
Nagilum...

Nagilum
Nagilum
Joined: 13 Feb 09
Posts: 12
Credit: 524868
RAC: 0

RE: It seems it may be a

Message 95751 in response to message 95749

Quote:
It seems it may be a problem with insufficient graphics memory at the time the problem occurred.


Well I decided that the CUDA functions for what ever reason, changed somehow, even though my graphics card is a Nvidia 9800 GX2 ,and now it was the card, that was the problem. After loading the Win 7 64bit version, I noticed some other compatibility problems. I decided to replace my Q6600 proccessor with the Q9650 chip and reinstall Win7 64bit. I've also disabled the CUDA in my preferances, and now the crunching has resumed flawlessly as before with twice the results as before. Somewhere along the line to improvement I have restored my reliability and doubled my production for Einstein@home. Also some of the games I play are now at maximum performance. I haven't told my wife how much I spent on the upgrade yet, but shes very accepting to my CPU goals and I'll soften the blow somehow. Thanks for the support and rapid responce to my inquiry. 64bit seems to be the way of the future!
Nagilum...

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5877
Credit: 118667223786
RAC: 19179148

RE: .... and now the

Message 95752 in response to message 95751

Quote:
.... and now the crunching has resumed flawlessly as before with twice the results as before. Somewhere along the line to improvement I have restored my reliability and doubled my production for Einstein@home.


Congratulations on solving your problems and your speed increase - well done!

Hope everything continues to go well for you.

Cheers,
Gary.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.