| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Discovering Dengue Drugs - Together
Thread: DDDT Single Redundancy Change |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 87
|
|
| Author |
|
|
mreuter80
Advanced Cruncher Joined: Oct 2, 2006 Post Count: 83 Status: Offline Project Badges: 
|
is it possible that an invalid WU turn valid at a later point (e.g. the 2nd computer returns a similar result?
|
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I'd say, very improbable. Once invalid, it already had a second opinion and is not considered again unless the validator is deemed faulty and manual intervention is required.
----------------------------------------So far i see about 1 in 10 on my machine going through the single distribution, being picked to get a second opinion (thus from "In Progress to "Inconclusive") and validate when that's done. A 100% score till now on all machines, so the interests is waning. At the outset knreed said to expect 1.2 results on average to get 1. Before it was 2.2 with redundancy. Logically in ZR you'd want to apply an initial higher safety, thus instead of 10%, 20% and still having won 1 result less, a huge improvement. The techs will tweak if they see it warranted to reduce the Random check, but why play it on the knifes edge if what we want is 100% reliable results for the scientists, with confidence. No one is reporting any anomalous "invalid", so we'll see.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Aug 12, 2008 11:09:47 AM] |
||
|
|
mreuter80
Advanced Cruncher Joined: Oct 2, 2006 Post Count: 83 Status: Offline Project Badges:
|
Hi Sekerob,
Thanks for the info. In my case the WU is invalid although the 2nd opinion is not returned, yet. dddt0602j0578_ 100507_ 1-- In Progress 08/12/2008 01:38:57 08/15/2008 02:06:33 0.00 0.0 / 0.0 dddt0602j0578_ 100507_ 0-- Invalid 08/09/2008 14:21:02 08/12/2008 00:44:17 8.04 103.0 / 0.0 Shouldn't it be of status "inconclusive" at this stage of validation? thanks |
||
|
|
mreuter80
Advanced Cruncher Joined: Oct 2, 2006 Post Count: 83 Status: Offline Project Badges:
|
nevermind, I found the answer in knreed's message:
"When a single redundancy result is returned, the mini workunit needs to match up with the value computed during the beta test (otherwise the result is marked invalid and the workunit is sent to someone else)." now it makes sense. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
(I now see the suggestion in your original question)
----------------------------------------Novel to me and wonder what this brought on. Possibly the mini wu that is run up front failed. knreed wrote in his point 4 what happens if the test is wrong but not how it would be marked in the interim: 4) We have also added some additional checks within the research apps to detect errant results. Part of this is a short run of the application that computes a known result (this was part of what we ran last week). This short run will be used to help ensure that the computation ran correctly as well as it will be used to determine the appropriate credit to award. So, please keep us posted on what it turns out to be when the second result returns. There is a good chance that only at that time the 50% credit is awarded as per the "invalid" rule. There is currently no valid reference to draw that factor from. Note to self: Update FAQ once we know the outcome as reported by mreuter80 [Edit: Now see a second post, finding some verbiage.... so it's indeed the failing of the mini WU by the looks. Still would like to cover it in the FAQ's once the second result is back)
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 2 times, last edit by Sekerob at Aug 12, 2008 3:46:11 PM] |
||
|
|
mreuter80
Advanced Cruncher Joined: Oct 2, 2006 Post Count: 83 Status: Offline Project Badges:
|
Hi Sekerob,
the status became now "valid" dddt0602j0578_ 100507_ 1-- Valid 08/12/2008 01:38:57 08/12/2008 22:59:22 12.09 96.3 / 99.6 dddt0602j0578_ 100507_ 0-- Valid 08/09/2008 14:21:02 08/12/2008 00:44:17 8.04 103.0 / 99.6 it seems the 2nd calculation confirmed the 1st answer and overruled the mini WU's result |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Good, a reverting to the classic quorum 2 validation and credit award accordingly. Amplified the FAQ for posterity.
----------------------------------------ciao
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Crystal Pellet
Veteran Cruncher Joined: May 21, 2008 Post Count: 1320 Status: Offline Project Badges: 
|
I had the same behaviour with at first an invalid result and afterwards it became valid because of the result of the second computer:
dddt0602m0770_ 100338_ 1-- In Progress 08/24/2008 10:37:55 08/26/2008 20:13:55 0.00 0.0 / 0.0 dddt0602m0770_ 100338_ 0-- Invalid 08/23/2008 11:51:38 08/24/2008 10:31:58 8.66 116.3 / 0.0 dddt0602m0770_ 100338_ 1-- Valid 08/24/2008 10:37:55 08/25/2008 01:20:56 5.04 96.5 / 106.4 dddt0602m0770_ 100338_ 0-- Valid 08/23/2008 11:51:38 08/24/2008 10:31:58 8.66 116.3 / 106.4 It seems that this mini prerun within the task should be overviewed. The redundancy is not so 'Single' as it perhaps could be, conceiving that most of the people will not notify and report those 'invalids'. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Opposed to 10% on quorum 2, it was said up front that 20% would be the redundancy on single distribution. 20% being the sum of repair work and random checking. If you see numbers higher than that on your client, posting the technical information (OS, BOINC version, CPU, RAM) would allow the other volunteers to draw comparisons and assess for themselves if it is a particular type of client doing this.
----------------------------------------[Added: WCG measures all the types of work going out to include the backup jobs for "invalid".]
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Aug 25, 2008 8:06:23 AM] |
||
|
|
Crystal Pellet
Veteran Cruncher Joined: May 21, 2008 Post Count: 1320 Status: Offline Project Badges:
|
Until now I'm under this 20%. Actually this was the first invalid one of 14 tasks, most of them on this machine (incl. the 'invalid' one):
Starting BOINC client version 6.2.16 for windows_x86_64 log flags: task, file_xfer, sched_ops Libraries: libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3 Running as a daemon Data directory: C:\ProgramData\BOINC Running under account boinc_master Processor: 4 AuthenticAMD AMD Phenom(tm) 9350e Quad-Core Processor [AMD64 Family 16 Model 2 Stepping 3] Processor features: fpu tsc pae nx sse sse2 pni OS: Microsoft Windows Vista: , Service Pack 1, (06.00.6001.00) Memory: 1.75 GB physical, 3.74 GB virtual Disk: 145.40 GB total, 123.98 GB free Local time is UTC +2 hours No coprocessors URL: http://abcathome.com/; Computer ID: 56793; location: home; project prefs: default URL: http://www.worldcommunitygrid.org/; Computer ID: 674419; location: (none); project prefs: default General prefs: from World Community Grid (last modified 23-May-2008 18:26:59) Host location: none General prefs: using your defaults Reading preferences override file Preferences limit memory usage when active to 1253.31MB Preferences limit memory usage when idle to 1611.40MB Preferences limit disk usage to 9.31GB Version 6.2.16  Sorry Sek, you're not happy.On two other machines running XP and WCG recommend version 5.10.45, but there I had only a few jobs until now. I was thinking this 20% redundancy was reserved for errors and no replies and not for valid 'ínvalids'. The next days I will only run DDDT besides ABC @ Home. If this behaviour happens more frequently I will report this here. |
|||
|
||||
|
|