| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: Inconclusive HPF2 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 19
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
It is back. But this time on a wu of HPF2. This can't be normal behaviour right?
----------------------------------------
[Edit 1 times, last edit by Former Member at May 18, 2007 9:12:51 PM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Don't know. The logic is that an 'attempt' is made after the first 15 are returned without error. If the statistical variation is too great, more copies are needed, to determine which one is not fitting the set, which is much tougher on HPF2 since all results are slightly different and come from 4 different operating system. With FAAH & GC the sets of 3 come from the same OS and are required to be the same.
----------------------------------------Can you sort on the 'return time' (hit the header label) and edit past into your original post, which helps reading the distribution sequence. Just had one validated, so lets wait what happens on the next few returns for your WU. lb868_ 00011_ 10-- Lapsed-01 Valid 05/15/2007 03:56:39 05/18/2007 17:33:28 11.04 118.4 / 117.9
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at May 18, 2007 9:04:58 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Sorted. This seems an isolated incident because other wu's of HPF2 are going fine as far as I can see. But this one looks rather strange though.
----------------------------------------Edit: I finally got it.  I thought inconclusive is given on individual basis but it is given to all the copies, except errors, as long as there are not enough reliable copies returned. So it is inconclusive or pending, both is not possible for copies of a wu. My posting is then a little bit early maybe to determine if anything is wrong, sorry about that.[Edit 3 times, last edit by Former Member at May 18, 2007 10:16:26 PM] |
||
|
|
coolbrook
Cruncher Joined: Mar 12, 2007 Post Count: 9 Status: Offline |
I have one of those too. Different WU. Hopefully it has a positive meaning aside from what's in the Wiki....
----------------------------------------Inconclusive Status Occasionally a workgroup returns 3 error free results before the deadline but the results do not match. In that case the 3 results are given Inconclusive status. Another copy of the WU is issued to another host with the hope that the result will match with 2 others and break the tie. Usually the 4th result does break the tie though on occassion a 5th or 6th WU must be issued to achieve a quorum. The 3 results that match are then given Valid status. Those that do not match receive Invalid status.   |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
A HPF2 job or another project i.e. FAAH or GC (FCG1)? Can you post the detail like Baron Samedi did in the first post, sorted like discussed in my first post in this thread).
----------------------------------------Sekerob PS, the WIKI does not discuss the entirely different HPF2 distribution which initially is 19, not 3! If you have one with 3 and a 4th shows, just sit back and relax.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The work unit in the opening post turns out to be fine. It's validated with one copy invalid. Cheers.
 |
||
|
|
olympic
Senior Cruncher Joined: Jun 12, 2005 Post Count: 156 Status: Offline |
I have 3 inconclusive HPF2's crunched by 3 different machines of mine.
----------------------------------------lb916_ 00030_ 16 lb909_ 00003_ 8 lb892_ 00044_ 15  |
||
|
|
coolbrook
Cruncher Joined: Mar 12, 2007 Post Count: 9 Status: Offline |
I didn't write the Wiki or see fit to edit it. Same idea.
----------------------------------------Mine is 19 lb924_00012 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Here's a little more information about validation, just to give you all a better idea of what's going on.
Each work unit has a minimum quorum - the number of results needed for validation. If, when you return your result, there aren't enough for validation, it is marked as "pending validation". When there are enough results pending, the validator can run. The validator for most projects (not HPF2) simply compares the results byte by byte. If all the results match, they are marked as "valid" and the work unit is complete. BOINC saves a copy of the valid result to compare future results to - so, if anybody sends a result in late (after successful validation) the result is simply compared to the canonical result, and marked "valid" or "invalid" accordingly. The late result isn't needed, and is ignored once it has been validated. If the results do not match, then there still aren't enough results for validation. The results are marked as "inconclusive" and extra copies are sent out to be crunched. In nearly all cases, just one of the results will actually be an error, and the others will match. So, once another result is in, the validator can run again and positively identify which is the error, and mark the others as valid. The recent FAAH errors were unusual in that there was a problem with the work. None of the results matched. When this happens, BOINC continues to send out copies until it reaches a limit. Once the limit is reached, all the results are marked as "error" and the work unit is reported to WCG as a problem unit. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thank you for explaining, Didactylos.
|
||
|
|
|