| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: The Clean Energy Project - Phase 2 Forum
Thread: CEP2 credit oddity!?!? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 11
|
|
| Author |
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Offline Project Badges: 
|
When checking up on a host that got an "out of the order" CEP2 WU, I noticed that it got a smaller than usual credit granted. When checking the WU Status, I saw this:
E208850_ 135_ C.28.C23H14N2SSeSi.01602513.3.set1d06_ 1-- 640 Valid 7/25/12 21:16:28 7/26/12 18:37:48 8.52 251.1 / 359.2On a multi-quorum WU, I would expect that the claimed credits are added up and the sum then evenly split between the reporting hosts. In this case however, the wingman got the full sum granted, while my host ended up getting less than a quarter of that sum...  Anyone care to explain? Ralf |
|||
|
||||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Are there any differences between the Result Logs of the two copies?
|
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Offline Project Badges:
|
Are there any differences between the Result Logs of the two copies? Come again?  Ralf |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
When checking up on a host that got an "out of the order" CEP2 WU, I noticed that it got a smaller than usual credit granted. When checking the WU Status, I saw this: E208850_ 135_ C.28.C23H14N2SSeSi.01602513.3.set1d06_ 1-- 640 Valid 7/25/12 21:16:28 7/26/12 18:37:48 8.52 251.1 / 359.2On a multi-quorum WU, I would expect that the claimed credits are added up and the sum then evenly split between the reporting hosts. In this case however, the wingman got the full sum granted, while my host ended up getting less than a quarter of that sum... Anyone care to explain? Ralf The result logs will tell the story. When both jobs do an equal amount of work then your *expectation* would be right. With the 12 hour bound reached, that rule does not apply, where the claim difference suggests at least a 2.5 times speed difference between the 2. Credit allocation principle is not different to HCMD2... number of jobs/attempts/seeds completed within 12 hours [for CEP2]. |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Offline Project Badges:
|
The result logs will tell the story. When both jobs do an equal amount of work then your *expectation* would be right. With the 12 hour bound reached, that rule does not apply, where the claim difference suggests at least a 2.5 times speed difference between the 2. Credit allocation principle is not different to HCMD2... number of jobs/attempts/seeds completed within 12 hours [for CEP2]. Well, how would I be able to see the result log of my wingman's WU?And I have certainly had WUs running longer than 12h and get "proper" credit for that. In several projects. In this case, the credit granting doesn't make any sense at all. Even if my slower host would get less credit as usual, why does the wingman get all the sudden the sum of both his and my credit claim? Ralf |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Offline Project Badges:
|
Here's an example of how I would expect the credit to be granted (this time my (different) hosts at the top, wingman at the bottom)
E208810_ 606_ C.29.C23H11N3S2Se.02056448.0.set1d06_ 1-- 640 Valid 7/23/12 08:43:30 7/24/12 01:41:40 9.27 149.8 / 133.5Claimed credit 149.8+117.3=267.1 Granted credit 267.1/2=133.5 Ralf |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
My Grid > Result Status > Click Work Unit name link and the status column links to every log of the results completed for that task.
You have not had tasks running longer than 12 hours *CPU* time. Elapsed time can run much longer, depends on how efficient your system is, but at the observed 2.5 times slower factor, doubt it is. Mental math says the sum is 359.0 not 359.2. It was of course not the credit rule applied [no such rule exists]. As said credit is computed based on work done in the maximum allowed time of 12 hours where the 8.52 hour task did it all [16 jobs], and the 12 hour task probably did not get passed job 3 or 4. |
||
|
|
branjo
Master Cruncher Slovakia Joined: Jun 29, 2012 Post Count: 1892 Status: Offline Project Badges: 
|
SR: in this project it is normal that "CPU time" and "elasped time" are not the same? Because I noticed this and thought it is a bug.
---------------------------------------- Crunching@Home since January 13 2000. Shrubbing@Home since January 5 2006  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello branjo,
Years ago BOINC switched over to report elapsed time (wall clock time) in an area that had previously reported CPU time. We still get queries about that change.  In general, you always have to remember that BOINC uses 2 different types of time and carefully distinguish between them. Lawrence |
||
|
|
branjo
Master Cruncher Slovakia Joined: Jun 29, 2012 Post Count: 1892 Status: Offline Project Badges:
|
Thanks Lawrence for info :-)
----------------------------------------Crunching@Home since January 13 2000. Shrubbing@Home since January 5 2006 |
||
|
|
|