| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Active Research
Forum: Mapping Cancer Markers Forum
Thread: MCM: Seeing interspersed Invalid results with 7.26, passing via PVal > PVerification |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 18
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I just noticed one there with a minimum quorum of 2 and replication of 5.
2 are marked valid and the other 3 are invalid, yet 4 of the 5 have the same value result.out. Mine is invalid and one of those 4. |
||
|
|
Mumak
Senior Cruncher Joined: Dec 7, 2012 Post Count: 477 Status: Offline Project Badges: 
|
I believe this is still the same issue with restoring checkpoints (after a machine restart, or BOINC exit) discussed in other threads. Those Invalids are most probably such machines.
---------------------------------------- |
||
|
|
Mumak
Senior Cruncher Joined: Dec 7, 2012 Post Count: 477 Status: Offline Project Badges:
|
The next Tuesday is the MS patch day, so a lot of reboots is going to happen.
----------------------------------------Techs, any hope this is going to be fixed ? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The next Tuesday is the MS patch day, so a lot of reboots is going to happen. Techs, any hope this is going to be fixed ? .....and as it is now the weekend we may not get a reply until Monday evening  |
||
|
|
ca05065
Senior Cruncher Joined: Dec 4, 2007 Post Count: 328 Status: Offline Project Badges: 
|
After several days of intermittent failures after overnight shutdown, I changed to my CEP2 method of snooze, wait one minute, sleep/hibernate, leave power connected overnight, switch on.
I have not had a Pval → Pver → invalid since then but I have noted odd behaviour which may help developers investigate the problem after the switch on. I have screen shots from part of Boinctasks window but the forum software does not allow me to post them. Time 00:19:28 - Normal running before starting overnight process: Time 00:20:37 - after snooze showing tasks suspended: Time 07:42:58 - after switch on which shows tasks running but CPU time not increasing: Time 07:43:04 - a few seconds later all tasks start a new checkpoint and continue normally: The output from stdoutdae.txt: 06-Dec-2013 23:51:42 [World Community Grid] Started download of MCM1_0000433_5716_MCM1_0000433_5716.txt 06-Dec-2013 23:51:44 [World Community Grid] Finished download of MCM1_0000433_5716_MCM1_0000433_5716.txt 07-Dec-2013 00:19:34 [---] Suspending computation - user request 07-Dec-2013 00:20:46 [---] Windows is suspending operations 07-Dec-2013 00:20:46 [---] Suspending network activity - requested by operating system 07-Dec-2013 00:20:54 [---] Resuming after OS suspension 07-Dec-2013 07:42:51 [---] Resuming computation 07-Dec-2013 07:42:51 [---] Resuming network activity 07-Dec-2013 07:42:52 [---] Windows is resuming operations 07-Dec-2013 07:53:50 [World Community Grid] Computation for task MCM1_0000423_4299_1 finished 07-Dec-2013 07:53:50 [World Community Grid] Starting task MCM1_0000427_7913_1 using mcm1 version 726 in slot 2 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Running the 7.26 32 bit version on my 64 bit Windows 7 system I got these errors running MCM1_0000429_0173_1 using mcm1 version 726:
12/7/2013 3:54:39 AM World Community Grid Started download of wcgrid_mcm1_graphics_prod_32.exe.7.26 12/7/2013 3:54:41 AM World Community Grid Finished download of wcgrid_mcm1_graphics_prod_32.exe.7.26 12/7/2013 7:52:08 AM Resuming computation 12/7/2013 8:01:58 AM World Community Grid Task MCM1_0000429_0173_1 exited with zero status but no 'finished' file 12/7/2013 8:01:58 AM World Community Grid If this happens repeatedly you may need to reset the project. 12/7/2013 8:01:59 AM World Community Grid Restarting task MCM1_0000429_0173_1 using mcm1 version 726 ... 12/7/2013 1:00:56 PM World Community Grid If this happens repeatedly you may need to reset the project. 12/7/2013 1:00:57 PM World Community Grid Restarting task MCM1_0000429_0173_1 using mcm1 version 726 12/7/2013 1:03:18 PM World Community Grid task MCM1_0000429_0173_1 aborted by user 12/7/2013 1:03:18 PM World Community Grid Starting MCM1_0000424_3039_0 12/7/2013 1:03:18 PM World Community Grid Starting task MCM1_0000424_3039_0 using mcm1 version 726 12/7/2013 1:03:20 PM World Community Grid Computation for task MCM1_0000429_0173_1 finished 12/7/2013 1:03:28 PM World Community Grid Resetting project |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
This is most often a sign of a busy machine. Make sure to check the Computing Preferences > Processor Usage > While processor usage is less than [or set it in the website device profiles]. Default on brand new member profiles is 50%. I'm running with 30% on Linux or else CEP2 jobs get this. Also in the Exclusive app tab [machine specific, not on web], you can specify programs that make BOINC pause when they are loaded [don't forget to exit them when done]. Again on my Linux I've got synaptic, clamtk, apt-get and apt-fast as exclusive apps. Many put their games in there... wow.exe or wot.exe [for Scribbles]. Then when running BOINC with "Run based on preferences", BOINC pauses [LAIM on, lossless suspending holding task in memory... eventually moved to VM if the other apps need more RAM]. Got a 100% success rate with these measures... zero invalids/error/restarts.
----------------------------------------edit: typo [Edit 1 times, last edit by Former Member at Dec 9, 2013 8:26:16 PM] |
||
|
|
armstrdj
Former World Community Grid Tech Joined: Oct 21, 2004 Post Count: 695 Status: Offline Project Badges: 
|
We are currently running a beta that addresses some of the invalids being seen in produciton but not all. We are able to reproduce the issue locally which is causing the remaining invalids so there will hopefully be another beta later this week with that fix. Please monitor the beta support forum for updates.
Thanks, armstrdj |
||
|
|
|