AP task completed in January still showing in results

Message boards : Number crunching : AP task completed in January still showing in results

To post messages, you must log in.

AuthorMessage
Profile Len
Avatar

Send message
Joined: 15 Mar 10
Posts: 53
Credit: 2,983,669
RAC: 1,569
United Kingdom
Message 1219626 - Posted: 17 Apr 2012, 18:20:21 UTC

This unit http://setiathome.berkeley.edu/workunit.php?wuid=905649963
was completed in January. My wingman completed it in February, and we have both apparently got credit. But it still shows up on my list.

Anyone know why?


I think I am. Therefore I am. I think.

ID: 1219626 · Report as offensive
Cruncher-American

Send message
Joined: 25 Mar 02
Posts: 1310
Credit: 175,917,184
RAC: 107,409
United States
Message 1219628 - Posted: 17 Apr 2012, 18:33:30 UTC - in response to Message 1219626.

This unit http://setiathome.berkeley.edu/workunit.php?wuid=905649963
was completed in January. My wingman completed it in February, and we have both apparently got credit. But it still shows up on my list.

Anyone know why?


No.

There appears to be a subtle code or database bug that allows this to happen - rarely. I have some around since last year.

ID: 1219628 · Report as offensive
MikeN

Send message
Joined: 24 Jan 11
Posts: 314
Credit: 44,958,927
RAC: 11,203
United Kingdom
Message 1219650 - Posted: 17 Apr 2012, 20:16:37 UTC - in response to Message 1219626.

This unit http://setiathome.berkeley.edu/workunit.php?wuid=905649963
was completed in January. My wingman completed it in February, and we have both apparently got credit. But it still shows up on my list.

Anyone know why?


This is a bug that was introduced around last August when they simplified the AP credit allocation code to solve some other problems. Essentially, the other two hosts had already been granted credit when you returned your result. You should still have got the same credit, but that subroutine appears to have been deleted. Judging by data of the server status page, there are around 13000 results in the same situation including 4 of mine, two of which have been waiting since last August. Hopefully once all remaining AP5's have been returned the guys in Berkeley will run a process to allocate the credit due to all of these

ID: 1219650 · Report as offensive
Les

Send message
Joined: 20 May 99
Posts: 53
Credit: 10,922,963
RAC: 8,784
United States
Message 1219669 - Posted: 17 Apr 2012, 20:59:07 UTC

The problem appears to be triggered when one computer fails to meet the assigned deadline but manages to report the result before the assigned replacement reports its result. Please look at the time stamps for the time reported and sent and you will see that computer 6212040 was sent the wu_2 version when computer 6108882 timed out. Then the next day computer 6108882 finally completed and reported the result for wu_1 and unfortunately was allowed to validate despite having work unit out in the field.

What seems to be happening is that the system is now waiting for the wingman to report to allow validation of the wu_2 result not realizing that there is no wingman assigned.

The previous version of this bug used to deny credit and validation to the computer that completed its work on time and instead validate the result from the delinquent computer including assignment of associated credit. This also means that canonical result may be assigned to the best analysis as that result may be the one that is being ignored.


ID: 1219669 · Report as offensive
Profile Wiggo "Socialist"
Avatar

Send message
Joined: 24 Jan 00
Posts: 10524
Credit: 135,315,346
RAC: 37,768
Australia
Message 1219670 - Posted: 17 Apr 2012, 21:07:24 UTC - in response to Message 1219669.

This problem should be fixed when the last remaining v505 tasks are done and returned (517 to go ATM) as the database should then be cleaned up before being put to rest.

Cheers.


ID: 1219670 · Report as offensive
David SProject Donor
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 17042
Credit: 20,939,152
RAC: 6,139
United States
Message 1219910 - Posted: 18 Apr 2012, 13:18:08 UTC - in response to Message 1219670.

This problem should be fixed when the last remaining v505 tasks are done and returned (517 to go ATM) as the database should then be cleaned up before being put to rest.

Cheers.

Just a couple days ago, I got a 505 that two previous hosts timed out on. I think I should make it.

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.


ID: 1219910 · Report as offensive
Profile Wiggo "Socialist"
Avatar

Send message
Joined: 24 Jan 00
Posts: 10524
Credit: 135,315,346
RAC: 37,768
Australia
Message 1219916 - Posted: 18 Apr 2012, 13:34:36 UTC - in response to Message 1219910.

This problem should be fixed when the last remaining v505 tasks are done and returned (517 to go ATM) as the database should then be cleaned up before being put to rest.

Cheers.

Just a couple days ago, I got a 505 that two previous hosts timed out on. I think I should make it.

436 left now at 13:20 UTC.

Cheers.

ID: 1219916 · Report as offensive

Message boards : Number crunching : AP task completed in January still showing in results


 
©2016 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.