Message boards :
Number crunching :
Odd Error Mesage in BOINC
Message board moderation
Author | Message |
---|---|
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
I just got about 300 or so of the following: 12/9/2010 9:07:35 AM SETI@home [error] No application found for task: windows_intelx86 610 cuda; discarding I think the servers think I am running Fermi???? And does "discarding" mean the WU was thrown away??? This is strange, because I am not running a Fermi card. Why would the servers try to send me such WUs? Also, from the name, is this an Intel-based app? My machines are all AMD, and have been forever.... |
Fred J. Verster Send message Joined: 21 Apr 04 Posts: 3252 Credit: 31,903,643 RAC: 0 |
Discarding means, your hosts aren't able to compute the tasks and thus sended to another host. Discard. Did you RESET SETI ? Or changed something in your BOINC config. Please, give some more info, BOINC version, optimized or stock application? Had a look at your hosts, OPTERON's and GTS240's, optimized, CUDA, anonymus, if I'm not mistaken :) I just received 16 CUDA WU's, after hitting the update button. I still run SSSE3x optimized and CUDA 6.08 on a (FERMI) GTX480, used the Lunatics Unified Installer 0.37 . |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
When I look at your account, it shows you have two computers attached. Both of them have CUDA capabilities. For some reason, BOINC could not find the application that processes CUDA work on your machine and aborted the work on your machine so that the servers can send that workunit out to someone else. As for the Intel vs. AMD, programmers like to name the app after the inventor of the architecture or instruction set. So you will always see Intel when standard x86 instructions are used. And if the programmer is being honest, you should always see AMD when x86-64 instructions are used. In your case, you are using a 32bit app of which Intel invented the 32bit instrcution set. Hence Intelx86. |
BilBg Send message Joined: 27 May 07 Posts: 3720 Credit: 9,385,827 RAC: 0 |
Post your app_info.xml using [pre] ... [/pre] tags (for easy reading - retains formatting/indent): [pre] ... {post contents of app_info.xml here} ... [/pre] I think you played with installing both the old v0.36 and the new v0.37 of the Lunatics' Unified Installer. Going from v0.36 to v0.37 is OK but not backwards (v0.37 -> v0.36) v0.37 can mark tasks as 6.08, 6.09 or 6.10 (all done by the same app so it doesn't matter what version the tasks are marked) but v0.36 marks CUDA tasks as only 6.08 so if you go back (v0.37 -> v0.36) all 6.09 or 6.10 CUDA tasks may be discarded (deleted). (v0.36 re-marks 6.09 to 6.08 but I don't remember will it re-mark 6.10 to 6.08) Also I see you are using the Fred's new BOINC Rescheduler: http://setiathome.berkeley.edu/forum_thread.php?id=60712 http://www.efmer.eu/forum_tt/index.php?topic=428.0 Do you use the last version (BoincRescheduler V 2.4)? (Or if you use the old ReSchedule 1.9 - don't! It is incompatible with Lunatics' Unified Installer v0.37) http://setiathome.berkeley.edu/forum_thread.php?id=54288&nowrap=true#1029244 Â - ALF - "Find out what you don't do well ..... then don't do it!" :) Â |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
I never "played" with going back to 0.36 from 0.37 (but I may have screwed up installing 0.37 on my machines...), and here's my app info: <app_info> <app> <name>setiathome_enhanced</name> </app> <file_info> <name>AK_v8b_win_x64_SSE3.exe</name> <executable/> </file_info> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>603</version_num> <platform>windows_intelx86</platform> <file_ref> <file_name>AK_v8b_win_x64_SSE3.exe</file_name> <main_program/> </file_ref> </app_version> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>603</version_num> <platform>windows_x86_64</platform> <file_ref> <file_name>AK_v8b_win_x64_SSE3.exe</file_name> <main_program/> </file_ref> </app_version> <app> <name>astropulse_v505</name> </app> <file_info> <name>ap_5.05r409_SSE.exe</name> <executable/> </file_info> <app_version> <app_name>astropulse_v505</app_name> <version_num>505</version_num> <platform>windows_intelx86</platform> <file_ref> <file_name>ap_5.05r409_SSE.exe</file_name> <main_program/> </file_ref> </app_version> <app_version> <app_name>astropulse_v505</app_name> <version_num>505</version_num> <platform>windows_x86_64</platform> <file_ref> <file_name>ap_5.05r409_SSE.exe</file_name> <main_program/> </file_ref> </app_version> <app> <name>setiathome_enhanced</name> </app> <file_info> <name>Lunatics_x32f_win32_cuda30_preview.exe</name> <executable/> </file_info> <file_info> <name>cudart32_30_14.dll</name> <executable/> </file_info> <file_info> <name>cufft32_30_14.dll</name> <executable/> </file_info> <file_info> <name>libfftw3f-3-1-1a_upx.dll</name> <executable/> </file_info> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>608</version_num> <platform>windows_intelx86</platform> <plan_class>cuda</plan_class> <avg_ncpus>0.040000</avg_ncpus> <max_ncpus>0.040000</max_ncpus> <coproc> <type>CUDA</type> <count>1</count> </coproc> <file_ref> <file_name>Lunatics_x32f_win32_cuda30_preview.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cudart32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>cufft32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>libfftw3f-3-1-1a_upx.dll</file_name> </file_ref> </app_version> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>609</version_num> <platform>windows_intelx86</platform> <plan_class>cuda23</plan_class> <avg_ncpus>0.040000</avg_ncpus> <max_ncpus>0.040000</max_ncpus> <coproc> <type>CUDA</type> <count>1</count> </coproc> <file_ref> <file_name>Lunatics_x32f_win32_cuda30_preview.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cudart32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>cufft32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>libfftw3f-3-1-1a_upx.dll</file_name> </file_ref> </app_version> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>610</version_num> <platform>windows_intelx86</platform> <plan_class>cuda_fermi</plan_class> <avg_ncpus>0.040000</avg_ncpus> <max_ncpus>0.040000</max_ncpus> <coproc> <type>CUDA</type> <count>1</count> </coproc> <file_ref> <file_name>Lunatics_x32f_win32_cuda30_preview.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cudart32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>cufft32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>libfftw3f-3-1-1a_upx.dll</file_name> </file_ref> </app_version> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>608</version_num> <platform>windows_x86_64</platform> <plan_class>cuda</plan_class> <avg_ncpus>0.040000</avg_ncpus> <max_ncpus>0.040000</max_ncpus> <coproc> <type>CUDA</type> <count>1</count> </coproc> <file_ref> <file_name>Lunatics_x32f_win32_cuda30_preview.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cudart32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>cufft32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>libfftw3f-3-1-1a_upx.dll</file_name> </file_ref> </app_version> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>609</version_num> <platform>windows_x86_64</platform> <plan_class>cuda23</plan_class> <avg_ncpus>0.040000</avg_ncpus> <max_ncpus>0.040000</max_ncpus> <coproc> <type>CUDA</type> <count>1</count> </coproc> <file_ref> <file_name>Lunatics_x32f_win32_cuda30_preview.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cudart32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>cufft32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>libfftw3f-3-1-1a_upx.dll</file_name> </file_ref> </app_version> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>610</version_num> <platform>windows_x86_64</platform> <plan_class>cuda_fermi</plan_class> <avg_ncpus>0.040000</avg_ncpus> <max_ncpus>0.040000</max_ncpus> <coproc> <type>CUDA</type> <count>1</count> </coproc> <file_ref> <file_name>Lunatics_x32f_win32_cuda30_preview.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cudart32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>cufft32_30_14.dll</file_name> </file_ref> <file_ref> <file_name>libfftw3f-3-1-1a_upx.dll</file_name> </file_ref> </app_version> </app_info> As far as the rescheduler is concerned, the XP64 system (unimatrix001) is using 2.4; the Vista-64 (BOINCBox) system is using 2.3. Also: does 0.37 install only a 32bit app? Because boh my machines are 64-bit... Thanks for your help... EDIT: Looking closer at the Meesages tab on BOINCBOX (where this occurred), i find that the error messages were followed by an approx. equal number of resends (265 of them!), so it looks like the servers fixed the problem for me)... but since the original error message doesn't specify what was dropped, I can't be certain. |
BilBg Send message Joined: 27 May 07 Posts: 3720 Credit: 9,385,827 RAC: 0 |
Can you post also the app_info.xml found in the oldApp_backup folder? (I hope it will be the version you used with v0.36 but this will be the case if you installed v0.37 only once) The original error message look strange: 12/9/2010 9:07:35 AM SETI@home [error] No application found for task: windows_intelx86 610 cuda; discarding It have to be: 12/9/2010 9:07:35 AM SETI@home [error] No application found for task: windows_intelx86 610 cuda_fermi; discarding _____ At first glance your current app_info.xml looks almost OK. The only thing I don't like is that you selected SSE3 Intel and not SSE3 AMD <name>AK_v8b_win_x64_SSE3.exe</name> It has to be <name>AK_v8b_win_SSE3_AMD.exe</name> To remedy - re-install the Win64 Lunatics' Unified Installer v0.37 and select SSE3 AMD (don't forget to select CUDA also) It doesn't matter is the app 32 or 64 bit, what matters is which is faster (and compatible). (if they include the 32 bit AK_v8b_win_SSE3_AMD.exe in the Win64 Installer this means they tested it and found this app to be faster for AMD CPUs even on 64 bit Windows. There is no such thing as "32 bit SSE" or "64 bit SSE" - only the integer instructions are 32 or 64 bit. ) Â - ALF - "Find out what you don't do well ..... then don't do it!" :) Â |
Miep Send message Joined: 23 Jul 99 Posts: 2412 Credit: 351,996 RAC: 0 |
your CPU app is 64bit so you picked the right installer ;) IIRC there is no 64bit GPU app because there was no advantage, but Jason will know for sure. Carola ------- I'm multilingual - I can misunderstand people in several languages! |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
APP_INFO from oldApps folder: <app_info> <app> <name>setiathome_enhanced</name> </app> <file_info> <name>AK_v8b_win_x64_SSE3.exe</name> <executable/> </file_info> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>528</version_num> <file_ref> <file_name>AK_v8b_win_x64_SSE3.exe</file_name> <main_program/> </file_ref> </app_version> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>603</version_num> <file_ref> <file_name>AK_v8b_win_x64_SSE3.exe</file_name> <main_program/> </file_ref> </app_version> <app> <name>astropulse_v5</name> </app> <file_info> <name>ap_5.03r112_SSE3.exe</name> <executable/> </file_info> <app_version> <app_name>astropulse_v5</app_name> <version_num>503</version_num> <file_ref> <file_name>ap_5.03r112_SSE3.exe</file_name> <main_program/> </file_ref> </app_version> <app> <name>astropulse_v505</name> </app> <file_info> <name>ap_5.05r168_SSE3.exe</name> <executable/> </file_info> <app_version> <app_name>astropulse_v505</app_name> <version_num>505</version_num> <file_ref> <file_name>ap_5.05r168_SSE3.exe</file_name> <main_program/> </file_ref> </app_version> <app> <name>setiathome_enhanced</name> </app> <file_info> <name>MB_6.08_CUDA_V12_VLARKill_FPLim2048.exe</name> <executable/> </file_info> <file_info> <name>cudart.dll</name> <executable/> </file_info> <file_info> <name>cufft.dll</name> <executable/> </file_info> <file_info> <name>libfftw3f-3-1-1a_upx.dll</name> <executable/> </file_info> <app_version> <app_name>setiathome_enhanced</app_name> <version_num>608</version_num> <plan_class>cuda</plan_class> <avg_ncpus>0.040000</avg_ncpus> <max_ncpus>0.040000</max_ncpus> <coproc> <type>CUDA</type> <count>1</count> </coproc> <file_ref> <file_name>MB_6.08_CUDA_V12_VLARKill_FPLim2048.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cudart.dll</file_name> </file_ref> <file_ref> <file_name>cufft.dll</file_name> </file_ref> <file_ref> <file_name>libfftw3f-3-1-1a_upx.dll</file_name> </file_ref> </app_version> </app_info> The same thing happened again just a short while ago - don't know what's happening here (on BOINCBox). I thought I did install the AMD version of the app - I certainly know I have Opteron processors! But, I may have spazzed when installing. I will do it again shortly. Thanks for your help! |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
Indeed. This query is likely to come up perodically so I'll explain. nVidia SDKs no longer support binding 32 bit device (GPU) code into 64 bit host executables (& vice versa), and 64 bit GPU code has considerable added overhead due to the larger pointers for addressing. It is only supported by Fermi anyway. One known special exception that would need the larger address space, would be a memory 'greedy' application built specifically to use all the memory on a 6GiB Tesla card. Since using RAM is the slowest part of Fermi, and we can process with a far smaller footprint, I'd have no intention of taking such an approach with current technologies. Jason "Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions. |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
Still getting the same behavior - right after BOINC startup, after the "app_info.xml file, using anonymous platform" message, I get a whole bunch of 12/9/2010 5:07:32 PM SETI@home [error] No application found for task: windows_intelx86 610 cuda; discarding and then a whole bunch of 12/9/2010 5:07:40 PM SETI@home Message from server: Resent lost task 06ja10af.26147.10292.13.10.93_1 and then the d/l of the files (20 at a time). So what IS going on? And why do the servers think I am running Fermi, when my machines have GT 240s? At this point, I am frustrated, but I can't do anything about it...if I'm not the only one having this problem, then an awful lot of b/w is being wasted. (I reinstalled Lunatics' 0.37, BTW). And this is still occurring... |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
Also, the BOINC Manager seems to be restarting every 3 hours, causing the same errors and results to occur...what the H is going on? |
arkayn Send message Joined: 14 May 99 Posts: 4438 Credit: 55,006,323 RAC: 0 |
|
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
Are you running one of the rescheduler's by chance? Yes - BoincRescheduler 2.3 on the machine in question. |
Gundolf Jahn Send message Joined: 19 Sep 00 Posts: 3184 Credit: 446,358 RAC: 0 |
Yes - BoincRescheduler 2.3 on the machine in question. That could explain the mismatch between your app_info.xml and the error message (there's no "windows_intelx86 610 cuda" combination in the file(s) you posted). Gruß, Gundolf |
BilBg Send message Joined: 27 May 07 Posts: 3720 Credit: 9,385,827 RAC: 0 |
What app is chosen here (in the red box): Choose 6.08, CUDA (I don't know what exactly is shown in this list - there have to be 3 choices in your case) Also, the BOINC Manager seems to be restarting every 3 hours, causing the same errors and results to occur...what the H is going on? You set BOINC Rescheduler to act automatically every 3 hours, check in the Settings tab. Â - ALF - "Find out what you don't do well ..... then don't do it!" :) Â |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
Thanks for your clues here.... 1) The app chosen is "none"; the 3 choices are: 6.08.cuda, 6.09.cuda23 and 6.10.cuda_fermi. Did I inadvertantly set some of my WUs to "6.10"? I don't know. If I did, how do I fix them, so this doesn't happen every time I shut down/restart BOINC? And what exactly is this feature used for? 2) The rescheduler is NOT set to run automatically. (I may have played with the setting mechanism when trying to learn how 2.3 works, but it was never started with the auto run timer set). And, according to the LOG tab, it hasn't been running every 3 hours...Actually, i checked BOINC's logs, and it HAS been running every 4 hours (not 3), or at least, BOINC has been restarting every 4 hours since sometime yesterday...does 2.3 keep a file somewhere that tells it to restart BOINC? or am I just screwed - should I reboot my machine to eliminate this 4-hour timer which mysteriously goes off DESPITE 2.3's settings???? It's "due" again in about an hour; I think I will see if it does restart again by itself.... |
S@NL - eFMer - efmer.com/boinc Send message Joined: 7 Jun 99 Posts: 512 Credit: 148,746,305 RAC: 0 |
I shouldn't matter what plan class You use for cuda work, they all point to the same exe: Lunatics_x32f_win32_cuda30_preview.exe. It is just a matter of taste if your card runs cuda_fermi or cuda. Check if these files are still there: Lunatics_x32f_win32_cuda30_preview.exe cudart32_30_14.dll cufft32_30_14.dll libfftw3f-3-1-1a_upx.dll BOINC is in the habit of just deleting any exe and dll he doesn't like. And when they are gone.... BOINC may discard all the work assigned to the GPU. And going back from V 0.37 is very very bad. And discarding is as bad as it sounds, gone..... TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking. |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
Fred - yup, all there, and I did NOT go backwards (.37 -> .36) Also: Yes, my BOINC was again restarted after 4 hours (9:07 EST) despite having no timer set in 2.3 Rescheduler; and no log entry in 2.3 either for the restart. So it looks like a 2.3 program bug of some kind. Later, since I was planning to move BOINCBox to a nice new HAF 932 case from its current crappy old Antec, I will of course shut down the system. That should kill the "phantom" timer. Stay tuned for more action!!! Meanwhile, I'll exit 2.3 and see if BOINC still restarts at 1:07... As far as Discarding goes - it would be nice if it told me exactly WHAT it wants to discard. Is it WUs? Which WUs? Or what other files, if any... |
BilBg Send message Joined: 27 May 07 Posts: 3720 Credit: 9,385,827 RAC: 0 |
It shouldn't matter what plan class You use for cuda work, they all point to the same exe: Lunatics_x32f_win32_cuda30_preview.exe. Yes - it doesn't matter but only if the tasks are marked (in client_state.xml) as one of these 3 combinations: 6.08.cuda 6.09.cuda23 6.10.cuda_fermi But why the discarded tasks are marked 6.10.cuda (impossible combination - and it is not normally found in app_info.xml) I don't know if the Fred's BOINC Rescheduler reads the info from app_info.xml or only from client_state.xml and where is the genesis of 6.10 + cuda ??? Â - ALF - "Find out what you don't do well ..... then don't do it!" :) Â |
S@NL - eFMer - efmer.com/boinc Send message Joined: 7 Jun 99 Posts: 512 Credit: 148,746,305 RAC: 0 |
Fred - yup, all there, and I did NOT go backwards (.37 -> .36) Just checked the code, but I can't find anything there. You could check and remove the check from "Automatic mode", just to be sure it's in a predictable state. But that should be impossible. TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking. |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.