| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Say No to Schistosoma Forum
Thread: Every WU result is Invalid |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 14
|
|
| Author |
|
|
pirogue
Veteran Cruncher USA Joined: Dec 8, 2008 Post Count: 685 Status: Offline Project Badges: 
|
I have one machine that is returning only Invalid WUs.
----------------------------------------From the status detail, it looks like the only differences between the valids and invalids are the time of day, CPU time, and BOINC version. All of the pdbqt size entries are the same. This particular machine is running 7.0.31. It keeps getting voted down by 6.10.X and 7.0.28 wingmen. Is there any way to tell why this is happening? edit: The device ID is 1664333. [Edit 1 times, last edit by pirogue at Nov 26, 2012 3:23:18 AM] |
||
|
|
keithhenry
Ace Cruncher Senile old farts of the world ....uh.....uh..... nevermind Joined: Nov 18, 2004 Post Count: 18667 Status: Offline Project Badges: 
|
If the detail from clicking on the "Invalid" link in RS doesn't show error errors or error codes, it may be that your machine is completing them too fast relative to the wingmen. What may be a valid result can get marked invalid IIFC if the claimed credit is too different from the wingman. I think in that case, it is sent to a third cruncher. If their claimed credit is more in line with the wingman, they and the wingman get valid and you get invalid. If it's OC'd, try it at stock for a bit and see if that makes a difference.
---------------------------------------- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
FAO of Techs:
Please confirm or reject: 1) The relative speed of devices in a quorum as impacting validation. 2) Credit difference having any impact on the rating of the validity of the result files. Thanks pirogue, *everything* going invalid is closest to there being an timing issue in the device... OC is one cause. Whilst the newer clients offer better management of GPU crunching, I can't possibly see the client doing anything to the science app - which is doing the work - 32 bit or 64 bit, Linux or Woz? |
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3315 Status: Offline Project Badges: 
|
I have one machine that is returning only Invalid WUs. From the status detail, it looks like the only differences between the valids and invalids are the time of day, CPU time, and BOINC version. All of the pdbqt size entries are the same. This particular machine is running 7.0.31. It keeps getting voted down by 6.10.X and 7.0.28 wingmen. Is there any way to tell why this is happening? edit: The device ID is 1664333. Which is the CPU?  - AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W - AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W - AMD Ryzen 7 7730U 8C/16T 3.0 GHz [Edit 1 times, last edit by Falconet at Nov 26, 2012 11:43:13 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Falconet, were you thinking of the sundry AMD CPU angle? Then the DSFL/GFAM would suffer the same faith [VINA jobs], the exception being that "Every" is being used as the operative word. With the [small group] of AMD inhomogeneity, some would validate if the device hits on the matching AMD wingman, not all. We'll know soon :|
|
||
|
|
branjo
Master Cruncher Slovakia Joined: Jun 29, 2012 Post Count: 1892 Status: Offline Project Badges: 
|
I am having the same problem with GFAM on 7.0.36 and 7.0.38(seems only when run via app info file).
---------------------------------------- Crunching@Home since January 13 2000. Shrubbing@Home since January 5 2006  |
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3315 Status: Offline Project Badges:
|
Falconet, were you thinking of the sundry AMD CPU angle? Then the DSFL/GFAM would suffer the same faith [VINA jobs], the exception being that "Every" is being used as the operative word. With the [small group] of AMD inhomogeneity, some would validate if the device hits on the matching AMD wingman, not all. We'll know soon :| Yes that was my idea. Maybe Pirogue isn't running those sciences and only knows of invalids in SN2S. We'll know soon. - AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W - AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W - AMD Ryzen 7 7730U 8C/16T 3.0 GHz |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
branjo may be onto something too. Vaguely recollect that phenomena of the app_info being mentioned before. The combined hardware/OS/Driver info may be needed as a reference and the science result history [His WCGDAWS will surely tell him in detail how that went down].
|
||
|
|
pirogue
Veteran Cruncher USA Joined: Dec 8, 2008 Post Count: 685 Status: Offline Project Badges:
|
It has dual Intel X5460s, is not overclocked, and not using an app_info.
----------------------------------------The completion times and credits claimed are not very different from the others. Examples: SN2S_ AAW88540_ 0000199_ 0037_ 2-- 620 Valid 11/25/12 08:23:22 11/26/12 03:25:13 5.27 102.9 / 106.4 <--6.10.58 SN2S_ AAW88540_ 0000199_ 0037_ 1-- 620 Valid 11/24/12 20:37:46 11/25/12 08:23:01 4.54 109.9 / 106.4 <--7.0.28 SN2S_ AAW88540_ 0000199_ 0037_ 0-- 620 Invalid 11/24/12 20:37:01 11/25/12 00:37:11 3.79 109.6 / 53.2 <--mine 7.0.31 SN2S_ AAW88540_ 0000177_ 0537_ 2-- 620 Valid 11/25/12 14:49:24 11/25/12 23:23:57 3.10 118.3 / 119.8 <--7.0.28 SN2S_ AAW88540_ 0000177_ 0537_ 1-- 620 Valid 11/24/12 12:43:22 11/25/12 14:48:10 5.21 121.2 / 119.8 <--6.10.17 SN2S_ AAW88540_ 0000177_ 0537_ 0-- 620 Invalid 11/24/12 12:42:56 11/24/12 16:46:54 4.01 115.2 / 59.9 <--mine 7.0.31 SN2S_ AAW88540_ 0000236_ 0063_ 2-- 620 Valid 11/26/12 03:29:54 11/26/12 14:58:48 5.17 140.4 / 142.8 <--6.12.33 SN2S_ AAW88540_ 0000236_ 0063_ 1-- 620 Valid 11/25/12 08:13:52 11/26/12 02:10:04 7.43 145.1 / 142.8 <-- 6.2.28 SN2S_ AAW88540_ 0000236_ 0063_ 0-- 620 Invalid 11/25/12 08:13:04 11/25/12 13:21:43 4.89 139.9 / 71.4 <--mine 7.0.31 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Now could that be an esoteric CPU? Not really, as BOINCStats reports there to be 387 devices at WCG i.e. the techs would have access to some comparative statistics, to single out incompatibility with wingmen of other CPU make.
|
||
|
|
|