rpi_logo
1.06 Errors with Windows Clients
1.06 Errors with Windows Clients
log in

Advanced search

Message boards : News : 1.06 Errors with Windows Clients

Author Message
Jeffery M. Thompson
Volunteer moderator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 23 Sep 12
Posts: 149
Credit: 12,948,617
RAC: 4,467

Message 57128 - Posted: 1 Feb 2013, 13:35:34 UTC
Last modified: 1 Feb 2013, 17:35:00 UTC

Windows 64 bit users seeing errors in work unit for Nbody Version 1.06 will want to use the reset option to have the software. This will redownload the dlls and fix a bad file that was transferred in the initial download.

Alternatively you can manually update the file per these steps
1. Stop MilkyWay@home
2. download the pthreadGC2_64_nbody_1.06.dll from http://milkyway.cs.rpi.edu/milkyway/download/pthreadGC2_64_nbody_1.06.dll
3. place pthreadGC2_64_nbody_1.06.dll into your mw directory


A more detailed discussion is in the following thread:

http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3128


Jeff

Profile BDDave
Avatar
Send message
Joined: 21 May 10
Posts: 18
Credit: 41,474,382
RAC: 41,715

Message 57131 - Posted: 1 Feb 2013, 14:46:02 UTC - in response to Message 57128.
Last modified: 1 Feb 2013, 14:49:25 UTC

Thanks Jeff,

Sounds like an easy fix, will reset project now.


Keep Crunchin'
Dave
____________

Richard Haselgrove
Send message
Joined: 4 Sep 12
Posts: 218
Credit: 448,778
RAC: 0

Message 57132 - Posted: 1 Feb 2013, 14:56:55 UTC - in response to Message 57128.

A more detailed discussion is in the following thread:

http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3128


Jeff

Making your link clickable.

Resetting the project is a safe and reliable way of overcoming this problem - though volunteers also running GPU applications for other parts of the milkyway project may wish to set 'No New Tasks' first and report all completed GPU work before resetting.

I'm a little worried by the 'download and rename' workround, for reasons I'll explain in the discussion thread, in a few minutes.

Jeffery M. Thompson
Volunteer moderator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 23 Sep 12
Posts: 149
Credit: 12,948,617
RAC: 4,467

Message 57136 - Posted: 1 Feb 2013, 17:36:54 UTC

Would it be fair to say the following?

So omitting the steps 4 and 5should suffice if reset can not be done.


Though using the reset is the preferred option.


Jeff

Richard Haselgrove
Send message
Joined: 4 Sep 12
Posts: 218
Credit: 448,778
RAC: 0

Message 57137 - Posted: 1 Feb 2013, 17:46:45 UTC - in response to Message 57136.

Placing the 'versioned' DLL (pthreadGC2_64_nbody_1.06.dll) in the project directory should be sufficient, because of the happy accident that the faulty version of the file and the replacement have exactly the same size.

If the two files had different sizes, BOINC would detect that the file had been "tampered with" at the next core client restart, and download a new copy anyway.

Thomas Weddige
Send message
Joined: 17 Feb 10
Posts: 1
Credit: 9,938,494
RAC: 1,821

Message 57156 - Posted: 4 Feb 2013, 9:14:48 UTC

Danke für die Info.
Mich hatte es schon gewundert warum den ganzen Tag andauernd Fehler produziert werden. ;-)

Und für alle die nur schlecht englisch können:

Milkyway@Home unter Projekte auswählen und auf Zurücksetzen klicken.
Dann wird die "Engine" neu geladen und der Fehler ist weg.
Aber vorher noch aktualisieren, damit die Punkte die noch hochgeladen werden müssen nicht weg sein. ;-)
____________

Nick Embrey
Send message
Joined: 4 Jul 09
Posts: 1
Credit: 2,333,863
RAC: 0

Message 57159 - Posted: 4 Feb 2013, 12:50:19 UTC

Might I suggest posting this message as a "notice" in the BOINC Manager? (Whatever that entails) Users aren't going to see it unless they go to milkyway's website, and it does seem like a lot of users haven't seen it yet, as many of my n-body tasks are still not granting credit due to too many errors by other users.

Mark, Sally, Jill, Laura and Craig
Send message
Joined: 28 Apr 10
Posts: 1
Credit: 8,933,269
RAC: 268

Message 57161 - Posted: 4 Feb 2013, 17:38:51 UTC

Anxiously applied this fix but am still having computation errors. Any other ideas would be greatly appreciated.

Thanks,

Mark

Len LE/GE
Send message
Joined: 8 Feb 08
Posts: 261
Credit: 104,050,322
RAC: 0

Message 57164 - Posted: 4 Feb 2013, 21:10:55 UTC - in response to Message 57161.

Anxiously applied this fix but am still having computation errors. Any other ideas would be greatly appreciated.


Seems the quick fix didn't work out for you.
Your choices are
- to set mw to NNT and after all tasks are sent back do a reset of the project
- or to disable n-body in the preferences until v1.07 is out (expected tomorrow)

In your case I would choose to reset mw to make sure you have a fresh and clean mw folder with no old/bad nbody dll hanging around.

Pornix
Send message
Joined: 16 Jul 10
Posts: 1
Credit: 2,272,828
RAC: 496

Message 57242 - Posted: 13 Feb 2013, 23:00:26 UTC
Last modified: 13 Feb 2013, 23:20:59 UTC

Hint me please!
Shall I let go and continue at the risk to sure go behind Validation Date with at least unit2 or skip them for renew? Why I ask? Well, you put the pressure on it, I got all the time it takes, when it rewards :)

http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=399687871(State:36%@43h)

http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=395028142(State:12%@90h)

Thanks for help in advance :)

Link
Avatar
Send message
Joined: 19 Jul 10
Posts: 327
Credit: 16,283,020
RAC: 0

Message 57249 - Posted: 14 Feb 2013, 8:33:00 UTC - in response to Message 57242.
Last modified: 14 Feb 2013, 8:34:09 UTC

Hint me please!
Shall I let go and continue at the risk to sure go behind Validation Date with at least unit2 or skip them for renew? Why I ask? Well, you put the pressure on it, I got all the time it takes, when it rewards :)

http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=399687871(State:36%@43h)

http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=395028142(State:12%@90h)

Thanks for help in advance :)

I'd say continue. It's highly unlikely, that the result that will be created and send out once your times out will be returned before yours. It might time out as well or even error out. Or become inconclusive and need another wingman anyway. So your result very likely will contribute to science and by not aborting it you also do not increase the probability, that these WUs get too many error results.
____________
.


Post to thread

Message boards : News : 1.06 Errors with Windows Clients


Main page · Your account · Message boards


Copyright © 2018 AstroInformatics Group