New NBody test searches

Author	Message
Steve Volunteer moderator Project developer Project tester Project scientist Send message Joined: 2 Jul 12 Posts: 4 Credit: 0 RAC: 0	Message 54980 - Posted: 3 Jul 2012 \| 20:37:14 UTC
Hello Everyone, My name is Steve, I am a physics/math undergrad at RPI and have taken on the task of handling the nbody code for the summer. For the past few weeks I have been working on the comparison software that measures how well the simulations match the known data. I just uploaded two searches, de_plum_slice_EMD_NEW_100K and ps_plum_slice_EMD_NEW_100K. Hopefully these will run smoothly but let me know if anyone has any problems. Other than that if anyone has any questions about me, what I do, or about Milky Way at Home, please feel free to ask. I would be happy to answer them.
ID: 54980 \| Rating: 0 \| rate: / Reply Quote

Penguin5540 Send message Joined: 4 Mar 12 Posts: 33 Credit: 16,760,223 RAC: 18,678	Message 54985 - Posted: 3 Jul 2012 \| 22:01:23 UTC
Getting computation errors on workunits named de_plum_slice_EMD_NEW_100K... all failing after 100% work done.
ID: 54985 \| Rating: 0 \| rate: / Reply Quote

Penguin5540 Send message Joined: 4 Mar 12 Posts: 33 Credit: 16,760,223 RAC: 18,678	Message 54986 - Posted: 3 Jul 2012 \| 22:03:33 UTC - in response to Message 54985.
Also on ps_plus_slice workunits. same thing. Computes then comes back with a computation error when at 100%
ID: 54986 \| Rating: 0 \| rate: / Reply Quote

Corsair Send message Joined: 13 Aug 09 Posts: 7 Credit: 24,198,443 RAC: 4,747	Message 54987 - Posted: 3 Jul 2012 \| 22:06:10 UTC
All workunits ps & ds crunched are getting error when reached 100% of computation. ____________ Corsair over the sailors grave never grows grass.
ID: 54987 \| Rating: 0 \| rate: / Reply Quote

Tag Send message Joined: 22 Mar 12 Posts: 1 Credit: 1,786,387 RAC: 324	Message 54989 - Posted: 3 Jul 2012 \| 23:03:29 UTC - in response to Message 54985.
Have just uploaded de_plum_slice_EMD_NEW_100K & ps_plum_slice_EMD_NEW_100K All have finished and ended with computation error. TAG
ID: 54989 \| Rating: 0 \| rate: / Reply Quote

JZD Send message Joined: 31 Dec 11 Posts: 3 Credit: 2,554,072 RAC: 7,335	Message 54990 - Posted: 4 Jul 2012 \| 0:07:11 UTC - in response to Message 54989. Last modified: 4 Jul 2012 \| 0:12:42 UTC
All have finished and ended with computation error. workunits http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=248916712, http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=248828315, http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=248917607
ID: 54990 \| Rating: 0 \| rate: / Reply Quote

Overtonesinger Send message Joined: 15 Feb 10 Posts: 52 Credit: 1,497,092 RAC: 1,281	Message 54995 - Posted: 4 Jul 2012 \| 7:05:47 UTC
All WUs from this set ended with computation error at 100 percent of completion: ps_plum_slice_EMD_NEW_100K_* example WU: http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=195811396 P.S. dear Steve, please, tell us a bit about Thyself! :) What is Thy greatest hobby? And Thy second hobby? (mine are: overtonesinging and computers). What is Thy current job or what does Thou study? :)
ID: 54995 \| Rating: 0 \| rate: / Reply Quote

Overtonesinger Send message Joined: 15 Feb 10 Posts: 52 Credit: 1,497,092 RAC: 1,281	Message 54996 - Posted: 4 Jul 2012 \| 7:27:17 UTC - in response to Message 54995.
http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=195808791 - WU name: de_plum_slice_EMD_NEW_100K_1341346802_15268 exit errorcode: <core_client_version>6.12.34</core_client_version> <![CDATA[ <message> System nemue nalezt uvedenou jednotku. (0xf) - exit code 15 (0xf) </message> <stderr_txt> <search_application> milkyway_nbody_0.80_windows_x86_64__mt.exe 0.80 Windows x86 double OpenMP, Crlibm </search_application> Using OpenMP 2 max threads on a system with 2 processors Could not load Ktm32.dll (1815): (null)Error reading histogram line 37: massPerParticle = 0.000100 09:14:29 (2428): called boinc_finish </stderr_txt> ]]> * "System nemuze nalezt uvedenou jednotku." is in Czech, so I think it is error-message of the Windows XP SP3 CZ running on that computer. - and it means: (Operating) System cannot find specified unit. (disk drive???) I guess there is some problem with a PATH to some file. :)[/code]
ID: 54996 \| Rating: 0 \| rate: / Reply Quote

Overtonesinger Send message Joined: 15 Feb 10 Posts: 52 Credit: 1,497,092 RAC: 1,281	Message 54998 - Posted: 4 Jul 2012 \| 9:54:28 UTC
Very interesting and weird Work-Unit! It has had veeeeery long run time for an NBody unit! :O how a pity it also errored out at 100 percent: http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=195808492
ID: 54998 \| Rating: 0 \| rate: / Reply Quote

TOTEM Send message Joined: 22 Jun 12 Posts: 2 Credit: 377,594 RAC: 0	Message 54999 - Posted: 4 Jul 2012 \| 10:11:19 UTC
MW@H 1.02 opencl_and n-body sim 0.80 showing computational error at 100%. Definitely weird.
ID: 54999 \| Rating: 0 \| rate: / Reply Quote

readingdancer Send message Joined: 11 May 11 Posts: 2 Credit: 1,991,722 RAC: 0	Message 55000 - Posted: 4 Jul 2012 \| 10:35:18 UTC
Hi Steve, Unfortunately these are also all producing a Computation error when they reach 100% for me too. Cheers, Chris
ID: 55000 \| Rating: 0 \| rate: / Reply Quote

Cartoonman Send message Joined: 10 Dec 09 Posts: 18 Credit: 6,518,358 RAC: 0	Message 55001 - Posted: 4 Jul 2012 \| 10:38:00 UTC
It appears, based on reviewing the stderr outputs of all of the users that have seen errors, that the application fails when reaching Histogram number 37 Error reading histogram line 37: massPerParticle = 0.000100 Another variant: Could not load Ktm32.dll (1815): (null)Error reading histogram line 37: massPerParticle = 0.000100
ID: 55001 \| Rating: 0 \| rate: / Reply Quote

greg_be Send message Joined: 18 Aug 09 Posts: 44 Credit: 953,427 RAC: 1,033	Message 55004 - Posted: 4 Jul 2012 \| 12:53:46 UTC - in response to Message 55001.
It appears, based on reviewing the stderr outputs of all of the users that have seen errors, that the application fails when reaching Histogram number 37 Error reading histogram line 37: massPerParticle = 0.000100 Another variant: Could not load Ktm32.dll (1815): (null)Error reading histogram line 37: massPerParticle = 0.000100 Another variant on the Ktm stuff: Could not load Ktm32.dll (126): The specified module could not be found.
ID: 55004 \| Rating: 0 \| rate: / Reply Quote

Saikrishna Send message Joined: 3 Apr 12 Posts: 1 Credit: 61,228 RAC: 277	Message 55005 - Posted: 4 Jul 2012 \| 12:56:37 UTC Last modified: 4 Jul 2012 \| 12:58:50 UTC
Same for me; all WUs related to plum_slice failed. Error was the histogram line 37 error, along with The system cannot find the drive specified. (0xf) - exit code 15 (0xf) http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=195678535 http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=195678534 http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=195678533 http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=195678532 http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=195673796 ____________
ID: 55005 \| Rating: 0 \| rate: / Reply Quote

TJ Send message Joined: 12 Aug 09 Posts: 251 Credit: 77,776,909 RAC: 957	Message 55006 - Posted: 4 Jul 2012 \| 13:32:16 UTC
Hi Steve, These "things": ps_plum_slice_EMD_NEW_100K_1341346802_25338, error out, not only at my place, also with wingman. So you have to study a wee bit more on the code... ____________ Greetings from, TJ
ID: 55006 \| Rating: 0 \| rate: / Reply Quote

Prroto Send message Joined: 21 Jun 12 Posts: 1 Credit: 1,122,762 RAC: 0	Message 55013 - Posted: 4 Jul 2012 \| 18:45:17 UTC Last modified: 4 Jul 2012 \| 18:45:46 UTC
Error, error, error... OMG! http://milkyway.cs.rpi.edu/milkyway/results.php?hostid=454572&offset=0&show_names=0&state=5&appid=
ID: 55013 \| Rating: 0 \| rate: / Reply Quote

Penguin5540 Send message Joined: 4 Mar 12 Posts: 33 Credit: 16,760,223 RAC: 18,678	Message 55014 - Posted: 4 Jul 2012 \| 21:24:05 UTC - in response to Message 55013.
errors continuing...
ID: 55014 \| Rating: 0 \| rate: / Reply Quote

Blurf Volunteer moderator Project administrator Send message Joined: 13 Mar 08 Posts: 617 Credit: 25,447,954 RAC: 0	Message 55017 - Posted: 4 Jul 2012 \| 21:36:29 UTC Last modified: 4 Jul 2012 \| 21:37:25 UTC
It's the 4th of July Holiday in the US....please don't expect any work to be done today. ____________
ID: 55017 \| Rating: 0 \| rate: / Reply Quote

Stephan Volkmann Send message Joined: 17 Mar 09 Posts: 7 Credit: 752,811 RAC: 1	Message 55019 - Posted: 4 Jul 2012 \| 23:30:48 UTC - in response to Message 55017.
Paket 195954817 Stephan Volkmann \| Abmelden Name de_plum_slice_EMD_NEW_100K_1341346802_24224 Anwendung MilkyWay@Home N-Body Simulation erstellt 4 Jul 2012 \| 12:34:45 UTC Mindestanzahl 1 Anfängliche Kopien 1 max # von Fehler/Gesamt/Erfolg Aufgaben 3, 9, 6 Fehler Zu viele Fehler (evtl. ein Bug) Aufgabe anklicken für Einzelheiten Computer Gesendet Meldezeit oder Ablaufdatum Erklärung Status Laufzeit (sek) CPU Zeit (sek) Punkte Anwendung 249287862 454178 4 Jul 2012 \| 12:40:07 UTC 4 Jul 2012 \| 21:23:38 UTC Fehler beim Berechnen 9.39 68.39 --- MilkyWay@Home N-Body Simulation v0.84 (mt) 249553467 450029 4 Jul 2012 \| 21:23:53 UTC 4 Jul 2012 \| 22:24:38 UTC Fehler beim Berechnen 8.03 76.51 --- MilkyWay@Home N-Body Simulation v0.88 (mt) 249584683 275574 4 Jul 2012 \| 22:24:43 UTC 4 Jul 2012 \| 23:25:16 UTC Fehler beim Berechnen 7.18 75.49 --- MilkyWay@Home N-Body Simulation v0.84 (mt) 249613873 454203 4 Jul 2012 \| 23:25:21 UTC 4 Jul 2012 \| 23:26:53 UTC Fehler beim Berechnen 19.04 33.69 --- MilkyWay@Home N-Body Simulation v0.88 (mt)
ID: 55019 \| Rating: 0 \| rate: / Reply Quote

Stephan Volkmann Send message Joined: 17 Mar 09 Posts: 7 Credit: 752,811 RAC: 1	Message 55020 - Posted: 4 Jul 2012 \| 23:32:07 UTC - in response to Message 55019.
248966335 439343 4 Jul 2012 \| 1:28:43 UTC 4 Jul 2012 \| 2:29:10 UTC Fehler beim Berechnen 256.37 909.62 --- MilkyWay@Home N-Body Simulation v0.88 (mt) 248999555 295489 4 Jul 2012 \| 2:35:31 UTC 4 Jul 2012 \| 3:45:32 UTC Fehler beim Berechnen 167.39 764.86 --- MilkyWay@Home N-Body Simulation v0.80 (mt) 249035935 432507 4 Jul 2012 \| 3:52:18 UTC 4 Jul 2012 \| 23:25:13 UTC Fehler beim Berechnen 302.74 1,114.24 --- MilkyWay@Home N-Body Simulation v0.80 (mt) 249613868 454203 4 Jul 2012 \| 23:25:21 UTC 4 Jul 2012 \| 23:29:39 UTC Fehler beim Berechnen 224.41 440.12 --- MilkyWay@Home N-Body Simulation v0.88 (mt)
ID: 55020 \| Rating: 0 \| rate: / Reply Quote

TimeRanger Send message Joined: 31 Oct 10 Posts: 19 Credit: 961,810 RAC: 1,499	Message 55023 - Posted: 5 Jul 2012 \| 9:22:25 UTC
6 of the EMD_NEW_100K units - 3 of the "de" and 3 "ps" .. all failed.
ID: 55023 \| Rating: 0 \| rate: / Reply Quote

Steve Volunteer moderator Project developer Project tester Project scientist Send message Joined: 2 Jul 12 Posts: 4 Credit: 0 RAC: 0	Message 55026 - Posted: 5 Jul 2012 \| 16:40:58 UTC Last modified: 5 Jul 2012 \| 16:49:09 UTC
Sorry for the late response time everyone, I was out of town yesterday for Forth of July shenanigans. It looks like the simulator is mostly running fine, but the program crashes at 100% when trying to load the new format of the output histograms. This should be a quick fix and we should have some proper simulations running soon. I'll post a message when I fix this problem and put up some more simulations. Thank you all for being patient. @Overtonesinger My greatest hobby would have to be skiing, it's pretty much all I do during the winter. Come the summertime I busy myself with rock climbing, juggling, and an online game call League of Legends. I would like to add piano playing to this list but sadly I don't really have access to one right now. My job for the summer is to get these simulations working again, and at RPI I study physics and mathematics of operations research.
ID: 55026 \| Rating: 0 \| rate: / Reply Quote

Overtonesinger Send message Joined: 15 Feb 10 Posts: 52 Credit: 1,497,092 RAC: 1,281	Message 55032 - Posted: 6 Jul 2012 \| 6:43:59 UTC
Steve has written: I just uploaded two searches, de_plum_slice_EMD_NEW_100K and ps_plum_slice_EMD_NEW_100K. Hopefully these will run smoothly but let me know if anyone has any problems. Dear Steve, well, as You see, they run pretty smuuuuthly. :))) Now seriously: It has been three days. Please, tell us about some progress, we deserve it. :) I have 3 small questions: 1) Can You estimate when they are fixed or deleted? 2) Because they replicate over and over when error, can You please set lower limit of their MAX.replication rate? ... So the last one of them will have their LAST TRY error out sooner ? :) 3) Tel us what to do with them. Compute untill they all error out their last try? Or shall we kill them on sight? ... shall we wait and compute something else for the time being untill fixed? 4) Can You please create new set of WUs that will work - to compute between the returning error-ones (there is already not enough work for all) ?? I have suggestion - as a professional programmer/developer: Now that You know for sure what is wrong with them(NEW) from all the posts, I would recommend: to compare the config of the "output generator" of the old ones and config of new ones, make it perfectly similar (search for hidden whitespace also! no extra tab or space should be there around the Line 37, just for sure ) - so it would be really the same format/syntax as the old ones, thus would run OK like the old WUs. Steve, could You try it, please? We need some new work for CPU (only NBody is really for CPU - the separation is done 800 times faster on GPUs... so better may it run on them). Thanx a lot for all answers to come, Steve. Namaste Filip
ID: 55032 \| Rating: 0 \| rate: / Reply Quote

Overtonesinger Send message Joined: 15 Feb 10 Posts: 52 Credit: 1,497,092 RAC: 1,281	Message 55033 - Posted: 6 Jul 2012 \| 6:57:23 UTC
Why not to use, just for now, the old format of the output histograms ?
ID: 55033 \| Rating: 0 \| rate: / Reply Quote

POPSIE Send message Joined: 25 Jan 11 Posts: 12 Credit: 3,643,137 RAC: 0	Message 55085 - Posted: 11 Jul 2012 \| 18:21:20 UTC Last modified: 11 Jul 2012 \| 18:27:30 UTC
After Restarting NBody-Server after last stop on beginning of month 6, I get many WU's. First 3 Day's all work well. On Third day, all WU's ending with calculation errors, I think ERR-NR15 After 4. day NBody Server where stopped and nobody gets new WU's Now no WU's and no running NBody-Server is the Staus of today. Over 6 weeks and no proper working NBody WOW!? <core_client_version>6.12.34</core_client_version> <![CDATA[ <message> process exited with code 15 (0xf, -241) </message> <stderr_txt> <search_application> milkyway_nbody 0.88 Linux x86_64 double OpenMP, Crlibm </search_application> Using OpenMP 7 max threads on a system with 4 processors Error reading histogram line 37: massPerParticle = 0.000100 18:53:54 (5221): called boinc_finish </stderr_txt>
ID: 55085 \| Rating: 0 \| rate: / Reply Quote

[AF>france>pas-de-calais]symaski62 Send message Joined: 7 Jul 09 Posts: 11 Credit: 4,406 RAC: 0	Message 55086 - Posted: 11 Jul 2012 \| 20:04:02 UTC Last modified: 11 Jul 2012 \| 20:19:13 UTC
http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=253993090 2.5% by 2.5% :)
ID: 55086 \| Rating: 0 \| rate: / Reply Quote

TLSI2000 Send message Joined: 15 Mar 10 Posts: 15 Credit: 72,106,380 RAC: 170,416	Message 55327 - Posted: 10 Aug 2012 \| 22:43:32 UTC
When the outstanding n-Body work units finally get down to zero, is there to be another series ? The count is now at 2.
ID: 55327 \| Rating: 0 \| rate: / Reply Quote

Author

Message

Steve
Volunteer moderator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jul 12
Posts: 4
Credit: 0
RAC: 0

Hello Everyone,

My name is Steve, I am a physics/math undergrad at RPI and have taken on the task of handling the nbody code for the summer. For the past few weeks I have been working on the comparison software that measures how well the simulations match the known data.

I just uploaded two searches, de_plum_slice_EMD_NEW_100K and ps_plum_slice_EMD_NEW_100K. Hopefully these will run smoothly but let me know if anyone has any problems.

Other than that if anyone has any questions about me, what I do, or about Milky Way at Home, please feel free to ask. I would be happy to answer them.

ID: 54980 | Rating: 0 | rate: