1.10 Error Thread
log in

Advanced search

Message boards : Number crunching : 1.10 Error Thread

1 · 2 · 3 · 4 · Next
Author Message
Profile nenym
Send message
Joined: 21 Jul 09
Posts: 11
Credit: 779,363,291
RAC: 232,418
Message 189 - Posted: 28 Jul 2009, 2:52:23 UTC

1.10 Tasks on Win XP x64 errored out. Host ID 812.
collatz_1248745476_537_1
collatz_1248745476_506_1
collatz_1248745476_322_1

example:
<core_client_version>6.6.28</core_client_version>
<![CDATA[
<stderr_txt>
Beginning processing...
Elapsed time: 0 seconds
called boinc_finish

</stderr_txt>
<message>
<file_xfer_error>
<file_name>collatz_1248745476_537_1_0</file_name>
<error_code>-161</error_code>
</file_xfer_error>

</message>
]]>

Profile kevint
Send message
Joined: 18 Jun 09
Posts: 34
Credit: 246,186,607
RAC: 0
Message 190 - Posted: 28 Jul 2009, 2:55:26 UTC - in response to Message 189.
Last modified: 28 Jul 2009, 3:03:29 UTC

Ditto: Win-64-CPU

host

7/27/2009 8:54:08 PM|Collatz Conjecture|Output file collatz_1248745476_1063_0_0 for task collatz_1248745476_1063_0 absent
7/27/2009 8:54:08 PM|Collatz Conjecture|Computation for task collatz_1248745476_1064_0 finished
7/27/2009 8:54:08 PM|Collatz Conjecture|Output file collatz_1248745476_1064_0_0 for task collatz_1248745476_1064_0 absent
7/27/2009 8:54:08 PM|Collatz Conjecture|Starting collatz_1248745476_952_0
7/27/2009 8:54:08 PM|Collatz Conjecture|Starting task collatz_1248745476_952_0 using collatz version 110
7/27/2009 8:54:10 PM|Collatz Conjecture|Computation for task collatz_1248745476_952_0 finished
7/27/2009 8:54:10 PM|Collatz Conjecture|Output file collatz_1248745476_952_0_0 for task collatz_1248745476_952_0 absent

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 1
Message 193 - Posted: 28 Jul 2009, 3:48:36 UTC - in response to Message 190.

I rebuilt the Win 64 bit app and set the version to v1.11
The compiler directive was messed up which controls whether to include the CUDA, ATI, or CPU code so it wasn't including any which is why it finished so fast without producing any output. Sorry about that.

frankhagen
Send message
Joined: 12 Jul 09
Posts: 188
Credit: 14,214,736
RAC: 1,406
Message 194 - Posted: 28 Jul 2009, 3:54:53 UTC

looks like the CUDA-version is running fine, but what's up with the validator?

30836 13279 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:26:10 UTC Completed, waiting for validation 0.87 0.01 pending
30834 13278 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:19:50 UTC Completed, waiting for validation 1.28 0.01 pending
30832 13277 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:12:28 UTC Completed, waiting for validation 1.59 0.01 pending
30830 13276 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:05:57 UTC Completed, waiting for validation 0.66 0.00 pending
30771 13246 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:34:28 UTC Completed, waiting for validation 0.95 0.01 pending
30769 13245 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:34:28 UTC Completed, waiting for validation 0.98 0.01 pending
30767 13244 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:26:10 UTC Completed, waiting for validation 1.44 0.01 pending
30765 13243 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:19:50 UTC Completed, waiting for validation 1.11 0.01 pending
30763 13242 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:12:28 UTC Completed, waiting for validation 1.36 0.01 pending
30761 13241 28 Jul 2009 2:12:06 UTC 28 Jul 2009 3:05:57 UTC Completed, waiting for validation 0.72 0.01 pending

____________

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 1
Message 196 - Posted: 28 Jul 2009, 4:06:15 UTC - in response to Message 194.

The work queue was essentially empty, so the first people done with a WU have to wait for their wingmen to finish before credit will be awarded (quorum of 2). I've been keeping an eye on the validator watching for a quorum to make sure everything is working OK and so far, so good.

Profile nenym
Send message
Joined: 21 Jul 09
Posts: 11
Credit: 779,363,291
RAC: 232,418
Message 197 - Posted: 28 Jul 2009, 4:11:07 UTC - in response to Message 193.
Last modified: 28 Jul 2009, 4:16:51 UTC

1.11 @ XP x64 runs OK, nice.

Thnx.

EDIT:
10% 7min32s
XEON 2.83GHz

frankhagen
Send message
Joined: 12 Jul 09
Posts: 188
Credit: 14,214,736
RAC: 1,406
Message 198 - Posted: 28 Jul 2009, 4:14:19 UTC - in response to Message 196.

The work queue was essentially empty, so the first people done with a WU have to wait for their wingmen to finish before credit will be awarded (quorum of 2). I've been keeping an eye on the validator watching for a quorum to make sure everything is working OK and so far, so good.


arrgh - it's too early over here - need more coffee.. ;)

just got the first one on my XEON running win64 - runtime prediction: 66 hours, boinc enters panic-mode. actual progess is more like 1%/minute..
____________

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 1
Message 199 - Posted: 28 Jul 2009, 4:29:39 UTC - in response to Message 198.

The work queue was essentially empty, so the first people done with a WU have to wait for their wingmen to finish before credit will be awarded (quorum of 2). I've been keeping an eye on the validator watching for a quorum to make sure everything is working OK and so far, so good.


arrgh - it's too early over here - need more coffee.. ;)

just got the first one on my XEON running win64 - runtime prediction: 66 hours, boinc enters panic-mode. actual progess is more like 1%/minute..


I haven't adjusted the flops yet for the new app. As soon as I get a few readings from new WUs completed on CPUs, I'll adjust the app's flop estimate accordingly.

Profile TomaszPawel
Avatar
Send message
Joined: 13 Jul 09
Posts: 29
Credit: 23,946,954
RAC: 0
Message 205 - Posted: 28 Jul 2009, 10:51:18 UTC - in response to Message 199.
Last modified: 28 Jul 2009, 10:52:33 UTC

Error to 1.10 CUDA 32bit:

When aplication is suspended, and resume it starts crunching WU from the begining.
(EG 95% -> suspend -> resume -> starts from 0%) sometimes it's happend sometimes not....
____________
POLISH NATIONAL TEAM - Join! Crunch! Win!

UBT - Ben
Send message
Joined: 26 Jul 09
Posts: 1
Credit: 1,214,330
RAC: 0
Message 207 - Posted: 28 Jul 2009, 11:08:40 UTC

Hi,

I downloaded the first couple of WU's. Stopped BOINC and installed the optimized ATI application.

All was going well until it hit about 3% and then the driver crashed and wouldn't work until i stopped and restarted BOINC. I have now dropped down to the SSE2 optimized app which is running fine so far, but i would much rather use the ATI application. :P

Windows error reporting service managed to get me some data in relation to the crash which is as follows:


Description
A problem with your video hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
OS Version: 6.0.6001.2.1.0.768.3
Locale ID: 2057

Files that help describe the problem
WD-20090728-1149.dmp
sysdata.xml
Version.txt

View a temporary copy of these files
Warning: If a virus or other security threat caused the problem, opening a copy of the files could harm your computer.

Extra information about the problem
BCCode: 117
BCP1: FFFFFA8004A48010
BCP2: FFFFFA6003416AC8
BCP3: 0000000000000000
BCP4: 0000000000000000
OS Version: 6_0_6001
Service Pack: 1_0
Product: 768_1


I am running Vista x64, with Catalyst 9.5 which has been running fine on MW. (I did suspend Milkyway first so it didn't interfere with this project).

Any ideas? :S

Thanks

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 1
Message 214 - Posted: 28 Jul 2009, 12:23:05 UTC - in response to Message 205.

Error to 1.10 CUDA 32bit:

When aplication is suspended, and resume it starts crunching WU from the begining.
(EG 95% -> suspend -> resume -> starts from 0%) sometimes it's happend sometimes not....


Could you please post a link to the WU to which you are referring?

Profile kevint
Send message
Joined: 18 Jun 09
Posts: 34
Credit: 246,186,607
RAC: 0
Message 218 - Posted: 28 Jul 2009, 13:37:00 UTC
Last modified: 28 Jul 2009, 13:43:08 UTC

Never mind...
Got it fixed

frankhagen
Send message
Joined: 12 Jul 09
Posts: 188
Credit: 14,214,736
RAC: 1,406
Message 221 - Posted: 28 Jul 2009, 14:07:56 UTC

any explanation why this happens?

32421 311 28 Jul 2009 2:55:44 UTC 28 Jul 2009 3:25:16 UTC Completed, marked as invalid 0.84 0.01 0.00
32422 652 28 Jul 2009 2:57:39 UTC 28 Jul 2009 3:08:34 UTC Completed and validated 50.29 0.30 0.00
34099 926 28 Jul 2009 4:15:45 UTC 28 Jul 2009 7:04:28 UTC Redundant result 0.00 --- ---
39742 458 28 Jul 2009 7:28:38 UTC 28 Jul 2009 8:52:00 UTC Completed and validated 3,526.28 21.11 0.00

____________

Profile Marty
Avatar
Send message
Joined: 12 Jul 09
Posts: 7
Credit: 269,978,547
RAC: 53
Message 222 - Posted: 28 Jul 2009, 14:24:39 UTC - in response to Message 221.

Tried to run the ATI/CAL version on XP32, Athlon XP 2400+ and HD3850 with Catalyst 8.12 installed and it errors out right away:

<core_client_version>6.5.0</core_client_version>
<![CDATA[
<message>
- exit code -1073741515 (0xc0000135)
</message>
]]>

WU
host

(Host runs the MW ATI/CAL program from Gipsel without errors)

Profile Gipsel
Volunteer moderator
Project developer
Project tester
Send message
Joined: 2 Jul 09
Posts: 279
Credit: 77,151,417
RAC: 77,866
Message 223 - Posted: 28 Jul 2009, 14:33:56 UTC - in response to Message 222.

Tried to run the ATI/CAL version on XP32, Athlon XP 2400+ and HD3850 with Catalyst 8.12 installed and it errors out right away:

<core_client_version>6.5.0</core_client_version>
<![CDATA[
<message>
- exit code -1073741515 (0xc0000135)
</message>
]]>

WU
host

(Host runs the MW ATI/CAL program from Gipsel without errors)

That's a typical error message when a dll file is missing.

Profile JerWA
Send message
Joined: 28 Jul 09
Posts: 57
Credit: 281,185,679
RAC: 5,179
Message 241 - Posted: 28 Jul 2009, 17:13:44 UTC
Last modified: 28 Jul 2009, 17:19:34 UTC

No error message that I noticed.

Win7 64 bit, ATI/CAL 1.10 app, HD4650, Catalyst 9.7, BOINC 6.6.36

Started GPU-Z 0.3.4, got a nasty long pause (not unexpected) and the Collatz WU hung. Said Running and timers were still going but no progress. Closed BOINC and restarted and the app fired back up from the last checkpoint (dropped 8% or so) and seems to be finishing fine.

GPU-Z and anything else that touches the drivers is known to cause issues, just thought I'd give you a heads-up that it seems to kill the app rather than causing the normal long pauses.

Edit: BOINC is also marking several running now, and with 2 active I get an almost immediate VPU crash, and on recovery all units still marked running and incrementing timers but not progressing. Changing app_info to -n1 appears to have fixed that.
____________

Profile [AF>Occitania>Lengadocian] F5LCU
Send message
Joined: 12 Jul 09
Posts: 6
Credit: 54,255,186
RAC: 0
Message 242 - Posted: 28 Jul 2009, 17:26:06 UTC

I tried v1.10 with boinc 6.6.36 on vista 32 bits.

Boinc launches v1.10 then vista told me the application has cease working.

All wus in error.

Seems not working here

Profile JerWA
Send message
Joined: 28 Jul 09
Posts: 57
Credit: 281,185,679
RAC: 5,179
Message 246 - Posted: 28 Jul 2009, 17:56:22 UTC

One of the WUs running when VPU crashed died, debugging info is in the err output:

http://boinc.thesonntags.com/collatz/result.php?resultid=59341

Profile Marty
Avatar
Send message
Joined: 12 Jul 09
Posts: 7
Credit: 269,978,547
RAC: 53
Message 250 - Posted: 28 Jul 2009, 18:41:02 UTC - in response to Message 223.

Tried to run the ATI/CAL version on XP32, Athlon XP 2400+ and HD3850 with Catalyst 8.12 installed and it errors out right away:

<core_client_version>6.5.0</core_client_version>
<![CDATA[
<message>
- exit code -1073741515 (0xc0000135)
</message>
]]>

WU
host

(Host runs the MW ATI/CAL program from Gipsel without errors)

That's a typical error message when a dll file is missing.

The amdcal*.dll's are in windows\system32 and the brook.dll is in the project folder. Any idea what else could be missing?

Is the file collatz_1.10_windows_intelx86__cal.exe maybe compiled with SSE2?
I was seeing a similar error on PG with a SSE2 application and the AXP only has SSE.

Matthias Lehmkuhl
Send message
Joined: 28 Jul 09
Posts: 6
Credit: 8,213,925
RAC: 3,115
Message 257 - Posted: 28 Jul 2009, 20:23:24 UTC

This result has crashed due the end of calculation
resultid=61821

<core_client_version>6.6.36</core_client_version>
<![CDATA[
<message>
Unzulässige Funktion. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
Beginning processing...
Collatz CUDA v1.10 (GPU Optimized Application)
worker: trying boinc_get_init_data()...
Looking for checkpoint file...
No checkpoint file found. Starting at beginning.
Success in SetCUDABlockingSync for device 0
CUDA Error: invalid device function
CUDA Kernel returned 0 steps
called boinc_finish

</stderr_txt>
]]>

Coprocessors NVIDIA GeForce 8600M GS (256MB) driver: 17948
____________
Matthias

1 · 2 · 3 · 4 · Next
Post to thread

Message boards : Number crunching : 1.10 Error Thread


Main page · Your account · Message boards


Copyright © 2018 Jon Sonntag; All rights reserved.