Astropulse Errors-Optimized version 5

Message boards : Number crunching : Astropulse Errors-Optimized version 5
Message board moderation

To post messages, you must log in.

1 · 2 · 3 · 4 . . . 6 · Next

AuthorMessage
Profile Blurf
Volunteer tester

Send message
Joined: 2 Sep 06
Posts: 8962
Credit: 12,678,685
RAC: 0
United States
Message 837034 - Posted: 4 Dec 2008, 21:42:02 UTC
Last modified: 21 Jan 2009, 21:48:55 UTC

This thread is to post errors and address concerns regarding the new Optimised AP v5.00 Astropulse application.

Please do not open new threads but post errors and commentary here.

Please, upgrade your optimized AP version to ap_5.00r103!

Thanks!


ID: 837034 · Report as offensive
Wild6-NJ
Volunteer tester

Send message
Joined: 4 Aug 99
Posts: 43
Credit: 100,336,791
RAC: 140
Message 837182 - Posted: 5 Dec 2008, 13:54:20 UTC
Last modified: 5 Dec 2008, 14:02:18 UTC

Greetings all-

THIS result was declared invalid and received zero credit, but the other two results for the wu got credit. Except for one other ap wu (also zero but waiting for a third result), all my others have received proper credit or are pending. Is there any reason for this?

(FYI: I may have upgraded to V5.00/4.37 while this wu was in process.)
ID: 837182 · Report as offensive
SlimDiesel

Send message
Joined: 2 Sep 99
Posts: 10
Credit: 15,496,641
RAC: 8
Canada
Message 837195 - Posted: 5 Dec 2008, 14:29:10 UTC - in response to Message 837182.  
Last modified: 5 Dec 2008, 14:31:59 UTC

Similar situation here? Several machines report this, mine among them (obviously) and the only one who got credit was the guy who aborted it???
ID: 837195 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 837196 - Posted: 5 Dec 2008, 14:55:38 UTC - in response to Message 837182.  

Greetings all-

THIS result was declared invalid and received zero credit, but the other two results for the wu got credit. Except for one other ap wu (also zero but waiting for a third result), all my others have received proper credit or are pending. Is there any reason for this?

(FYI: I may have upgraded to V5.00/4.37 while this wu was in process.)

Looks like a half way through switchover. I'd suggest to forget about this one.
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 837196 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 837197 - Posted: 5 Dec 2008, 14:57:10 UTC - in response to Message 837195.  

Similar situation here? Several machines report this, mine among them (obviously) and the only one who got credit was the guy who aborted it???


There is seemingly no optimised application represtented in any of the tasks here. Are you sure this relates to v5 Optimised

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 837197 · Report as offensive
| ganja |
Volunteer tester
Avatar

Send message
Joined: 9 Aug 08
Posts: 26
Credit: 2,463,915
RAC: 0
United States
Message 837342 - Posted: 6 Dec 2008, 1:33:54 UTC

I am running the opti v5 for astropulse and I have had two out of five give me 0 credit has anyone else had this problem the rest give me around 757 or so.

ganja
ID: 837342 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 837391 - Posted: 6 Dec 2008, 6:26:14 UTC - in response to Message 837342.  

I am running the opti v5 for astropulse and I have had two out of five give me 0 credit has anyone else had this problem the rest give me around 757 or so.

ganja

The "Valid" indication is bogus, neither WU 373524588 nor WU 351720292 have been decided yet. Many users have seen the same, just look to see if a canonical result has been chosen to know if final judgement has been reached.

WU 351720292 may be in danger of erroring out, so cross your fingers on that one.
                                                               Joe
ID: 837391 · Report as offensive
Profile Gustav_and_Padma
Avatar

Send message
Joined: 26 Oct 03
Posts: 16
Credit: 315,654
RAC: 0
United States
Message 837759 - Posted: 7 Dec 2008, 16:55:59 UTC - in response to Message 837391.  

What happens when a successfully completed WU waits over a month for a wingman to validate? Included in our pending credit total, we had an AP WU of about 750 credits pending last night. But this morning they were dropped from our pending credit total and the result is no longer visible. Our total credit didn't increase accordingly. Was the work lost?

This other one is a month old too, and we're wondering if results like these will eventually get considered, or if we are just wasting electricity.
http://setiathome.berkeley.edu/workunit.php?wuid=354147917
ID: 837759 · Report as offensive
Profile Leaps-from-Shadows
Volunteer tester
Avatar

Send message
Joined: 11 Aug 08
Posts: 323
Credit: 259,220
RAC: 0
United States
Message 837835 - Posted: 7 Dec 2008, 22:37:15 UTC

There is no time limit for Pending units. Work units will appear in Pending until they are validated, period. It doesn't matter whether they are Astropulse or Multibeam. I've had Pending work units for over nine weeks.

If a work unit disappeared from your Pending, it was validated. You may not have gotten credit for it because of the recent transition.

You are able to see the results for 24 hours after validation. Once those 24 hours are up, the visible info gets deleted.
Cruiser
Gateway GT5692 L-f-S Edition
-Phenom X4 9650 CPU
-4GB 667MHz DDR2 RAM
-500GB SATA HD
-Vista x64 SP1
-BOINC 6.2.19 32-bit client
-SSE3 optimized 32-bit apps
ID: 837835 · Report as offensive
Cosmic_Ocean
Avatar

Send message
Joined: 23 Dec 00
Posts: 3027
Credit: 13,516,867
RAC: 13
United States
Message 837836 - Posted: 7 Dec 2008, 22:42:06 UTC

What I have found with the AP units is if they go into the third-party verification state, the granted credit changes to zero until the third result comes in, and since some kind of credit was granted, it disappears from pending. I've got several APs that are like that.
Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)
ID: 837836 · Report as offensive
Profile Gustav_and_Padma
Avatar

Send message
Joined: 26 Oct 03
Posts: 16
Credit: 315,654
RAC: 0
United States
Message 837891 - Posted: 8 Dec 2008, 3:26:39 UTC - in response to Message 837835.  

OK. thanx. Actually since the last time we checked, our total credit just jumped up to where we assumed it would go if granted crdit. So, even tho some umtold number of pending AP WU's are still ver 4,(not just ours but everyone elses) they are likely to validate too, it would seem. We could be wrong about this: What we are saying is that we don't think the credit was visible for 24 hours after it disappeared from pending status. And it seems like it took well over 12 hours before the credit was added to our total. We are just glad that the older version AP WU's are able to validate at all. Cheers!
ID: 837891 · Report as offensive
mmonroe

Send message
Joined: 22 Oct 06
Posts: 20
Credit: 5,132,754
RAC: 0
United States
Message 838248 - Posted: 9 Dec 2008, 13:59:33 UTC

mmaybe this is another of these. I run 4.36; wingman runs 5.00



http://setiathome.berkeley.edu/workunit.php?wuid=360133998
ID: 838248 · Report as offensive
Profile Byron S Goodgame
Volunteer tester
Avatar

Send message
Joined: 16 Jan 06
Posts: 1145
Credit: 3,936,993
RAC: 0
United States
Message 838253 - Posted: 9 Dec 2008, 14:06:46 UTC - in response to Message 838248.  

mmaybe this is another of these. I run 4.36; wingman runs 5.00



http://setiathome.berkeley.edu/workunit.php?wuid=360133998

Yes it appears to be.
ID: 838253 · Report as offensive
mmonroe

Send message
Joined: 22 Oct 06
Posts: 20
Credit: 5,132,754
RAC: 0
United States
Message 838264 - Posted: 9 Dec 2008, 14:45:12 UTC - in response to Message 838253.  

[quote]mmaybe this is another of these. I run 4.36; wingman runs 5.00



http://setiathome.berkeley.edu/workunit.php?wuid=360133998

Yes it appears to be.[/quotegThanks what can I do about it, if anything?
ID: 838264 · Report as offensive
Profile Byron S Goodgame
Volunteer tester
Avatar

Send message
Joined: 16 Jan 06
Posts: 1145
Credit: 3,936,993
RAC: 0
United States
Message 838267 - Posted: 9 Dec 2008, 14:52:34 UTC - in response to Message 838264.  
Last modified: 9 Dec 2008, 14:53:57 UTC

mmaybe this is another of these. I run 4.36; wingman runs 5.00



http://setiathome.berkeley.edu/workunit.php?wuid=360133998

Yes it appears to be.
Thanks what can I do about it, if anything?

Not much you can do but wait for the wingman to complete the task so that the validity of the tasks and credit can be determined.
ID: 838267 · Report as offensive
mmonroe

Send message
Joined: 22 Oct 06
Posts: 20
Credit: 5,132,754
RAC: 0
United States
Message 838347 - Posted: 9 Dec 2008, 22:46:45 UTC - in response to Message 838267.  

mmaybe this is another of these. I run 4.36; wingman runs 5.00



http://setiathome.berkeley.edu/workunit.php?wuid=360133998

Yes it appears to be.
Thanks what can I do about it, if anything?

Not much you can do but wait for the wingman to complete the task so that the validity of the tasks and credit can be determined.


Thanks. A quick question:

I am using the app downloaded from SETI; I choose not to use the optimized app because I am concerned about the upkeep and my technical capability to install it correctly. I understand the main advantage of the app is that it runs ap faster.Given that, how is an app from outside the SETI group allowed to run and negate the results of the SETI app I run? It appears to me that whoever coded the optimized app failed to properly vet it against the stock app many of us use so that the results were similar and, more important accepatble to SETI.

How can this happen?
ID: 838347 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14650
Credit: 200,643,578
RAC: 874
United Kingdom
Message 838349 - Posted: 9 Dec 2008, 23:09:20 UTC - in response to Message 838347.  

Thanks. A quick question:

I am using the app downloaded from SETI; I choose not to use the optimized app because I am concerned about the upkeep and my technical capability to install it correctly. I understand the main advantage of the app is that it runs ap faster.Given that, how is an app from outside the SETI group allowed to run and negate the results of the SETI app I run? It appears to me that whoever coded the optimized app failed to properly vet it against the stock app many of us use so that the results were similar and, more important accepatble to SETI.

How can this happen?

Just running the stock app is fine - your contribution to SETI@Home is still welcomed and appreciated. Optimised apps are sometimes fiddly and time-consuming to manage: you've made an honest appraisal of the committment you're prepared to make, and come to an appropriate decision.

With regard to the validation problem: the project in fact does officially support and endorse the use of optimised applications - the source code is made publicly available for volunteers to work on, and many improvements over the years have been initiated by volunteers but subsequently accepted back into the 'official' applications. The optimisation programmers go to great lengths, as a matter of pride, to ensure that their work is a least as good as, and sometimes better, than the Berkeley applications, and that the results validate as reliably as possible.

In the case of your WU 1047643474, none of this applies: both you and your wingman used the stock app downloaded from SETI. Neither of you used an optimised application. The trouble is, the two of you calculated the results a month appart (somebody else neglected to return the task allocated to them), and in the meantime Berkeley updated their application - and the new version doesn't match the old one. Something went wrong with either the planning or the implementation of the transition: we don't know what, because neither of the programmers involved has posted on a public website since it happened. We can only speculate.
ID: 838349 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 838363 - Posted: 10 Dec 2008, 1:10:29 UTC - in response to Message 838347.  

mmaybe this is another of these. I run 4.36; wingman runs 5.00

http://setiathome.berkeley.edu/workunit.php?wuid=360133998


I am using the app downloaded from SETI; I choose not to use the optimized app because I am concerned about the upkeep and my technical capability to install it correctly. I understand the main advantage of the app is that it runs ap faster.Given that, how is an app from outside the SETI group allowed to run and negate the results of the SETI app I run? It appears to me that whoever coded the optimized app failed to properly vet it against the stock app many of us use so that the results were similar and, more important accepatble to SETI.

How can this happen?

The 5.00 optimized app being used by host 4009342 will probably do what it is designed to, and match the stock 5.00 result. If there's a problem with the optimized app it won't match and yet another host will be given the WU.

There are still many possible outcomes, here are some:

1) The next result matches the 5.00 stock result, your result from 4.36 isn't close enough, but Eric's credit granting script runs and gives you credit because you crunched in good faith.

2) The next result matches the 5.00 stock result, your 4.36 result is weakly similar so the Validator grants credit.

3) The original wingman's host finishes the WU before the new host, the result is strongly similar to yours, and yours becomes the canonical result.

4) The WU gets to the maximum of 10 results without a match and nobody gets credit.

5) The project shuts down for lack of funds and nobody gets any credits.

I consider 1) most likely, though there's certainly the possibility the script won't be run at the right time to help or it doesn't handle this specific situation. 2) is definitely possible though I can't judge how likely. 3), 4), and 5) are quite unlikely of course.
                                                              Joe
ID: 838363 · Report as offensive
mmonroe

Send message
Joined: 22 Oct 06
Posts: 20
Credit: 5,132,754
RAC: 0
United States
Message 838376 - Posted: 10 Dec 2008, 1:59:00 UTC - in response to Message 838349.  

Thanks. A quick question:

I am using the app downloaded from SETI; I choose not to use the optimized app because I am concerned about the upkeep and my technical capability to install it correctly. I understand the main advantage of the app is that it runs ap faster.Given that, how is an app from outside the SETI group allowed to run and negate the results of the SETI app I run? It appears to me that whoever coded the optimized app failed to properly vet it against the stock app many of us use so that the results were similar and, more important accepatble to SETI.

How can this happen?

Just running the stock app is fine - your contribution to SETI@Home is still welcomed and appreciated. Optimised apps are sometimes fiddly and time-consuming to manage: you've made an honest appraisal of the committment you're prepared to make, and come to an appropriate decision.

With regard to the validation problem: the project in fact does officially support and endorse the use of optimised applications - the source code is made publicly available for volunteers to work on, and many improvements over the years have been initiated by volunteers but subsequently accepted back into the 'official' applications. The optimisation programmers go to great lengths, as a matter of pride, to ensure that their work is a least as good as, and sometimes better, than the Berkeley applications, and that the results validate as reliably as possible.

In the case of your WU 1047643474, none of this applies: both you and your wingman used the stock app downloaded from SETI. Neither of you used an optimised application. The trouble is, the two of you calculated the results a month appart (somebody else neglected to return the task allocated to them), and in the meantime Berkeley updated their application - and the new version doesn't match the old one. Something went wrong with either the planning or the implementation of the transition: we don't know what, because neither of the programmers involved has posted on a public website since it happened. We can only speculate.

OK not to sound stupid but I am using the 6.2.19 app from SETI. How do I get the 5.00 app you refer to?
ID: 838376 · Report as offensive
Profile dnolan
Avatar

Send message
Joined: 30 Aug 01
Posts: 1228
Credit: 47,779,411
RAC: 32
United States
Message 838379 - Posted: 10 Dec 2008, 2:16:59 UTC - in response to Message 838376.  
Last modified: 10 Dec 2008, 2:19:58 UTC


OK not to sound stupid but I am using the 6.2.19 app from SETI. How do I get the 5.00 app you refer to?


6.2.19 is the version of Boinc, 5.00 is the version of the AP science app. If you are running stock with no app_info.xml file, you probably already have it,
[edit] or will when you get new Astropulse work [/edit] otherwise you have to download the appropriate package. You can check here for a download link.

-Dave
ID: 838379 · Report as offensive
1 · 2 · 3 · 4 . . . 6 · Next

Message boards : Number crunching : Astropulse Errors-Optimized version 5


 
©2024 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.