0 credits granted
17 Feb 2005 14:51:41 UTC
Topic 187680
(moderation:
I completed a WU, but was granted 0 credit. Two other computers finished this WU and were granted credit. Is this normal?
http://einsteinathome.org/workunit/325834
0 credits granted
)
> I completed a WU, but was granted 0 credit. Two other computers finished this
> WU and were granted credit. Is this normal?
It failed validation. :(
http://einsteinathome.org/task/1154022
Be lucky,
Neil
> I completed a WU, but was
)
> I completed a WU, but was granted 0 credit. Two other computers finished this
> WU and were granted credit. Is this normal?
Yes.
One of the "core", um, features of the BOINC concept is that there will be work that will not gain credit. This is spelled out in the "Rules and Policies" section of the Project. The bottome line is that in the bad old days project like SETI@Home gave credit where none was due. So, we spawned counterfiters, cheaters, gremlins, and hateful personalities ... :)
Seriously, we are doing science, and sometimes we don't get the answer we want. Other times we don't even get clear failures of the work. Anyway, we all have work that fails for one reason or another.
Hi ! Don't you think it
)
Hi !
Don't you think it would be better to send a "stop computing" signal to all others computers that you don't need results ?
1- never having message "Result was reported too late to validate"
2- like that, you and us gain time to get advance in research
This is not for credits problem, but only not to spend time for nothing
Am i right or not ?
Honi soit ki mal y pense !
> Hi ! > > Don't you think
)
> Hi !
>
> Don't you think it would be better to send a "stop computing" signal to all
> others computers that you don't need results ?
No way to do this. My computers at least are behind a firewall that is difficult to get through, and the addresses of the computers are non-routable from outside.
>
> 1- never having message "Result was reported too late to validate"
>
The BOINC client CPU scheduler and download scheduler should be rewritten to correctly deal with multiple projects.
> 2- like that, you and us gain time to get advance in research
>
> This is not for credits problem, but only not to spend time for nothing
>
> Am i right or not ?
>
BOINC WIKI
The Project Developers can't
)
The Project Developers can't do all our thinking for us. Here's a short list of TO DO's when participating in single or multiple BOINC project(s):
1. Be responsible. Adjust your project settings so as not to download more WU's than you can complete in the allotted time.
2. Complete the WU's you've downloaded.
It's a pretty short list.
There's a fine line between fishing and standing on the shore looking like an idiot -- Steven Wright
>The Project Developers can't
)
>The Project Developers can't do all our thinking for us. Here's a short list >of TO DO's when participating in single or multiple BOINC project(s):
>
>
>1. Be responsible. Adjust your project settings so as not to download more >WU's than you can complete in the allotted time.
>
>2. Complete the WU's you've downloaded.
>
>
>It's a pretty short list.
Sound advice for all BOINCers!!!
During the past couple of
)
During the past couple of weeks I've made a number of changes to the BOINC scheduler, specifically designed to help prevent machines from downloading more work than they can complete by the 'Report Deadline'. However as has been pointed out in this thread, there is no substitute for intelligent and responsible management by you and other users. Keep an eye on your work queue and deadlines, and try and set your project preferences, work caches, and resource shares in a responsible way. This helps to maximize your overall contribution to the project.
Cheers,
Bruce
Director, Einstein@Home
So how about this
)
So how about this one:
http://einsteinathome.org/workunit/347835
I returned the result in plenty of time. In fact, before one of the others returned their result, but where they were granted credit, my result was granted 0.0.
Browsing through the results, it seem this is not altogether uncommon.
I am assuming it is some kind of validation failure. The other two agreed and mine did not agrree with them. Could it have something to do with the fact that my result was computed under Linux and the others were computed under Windows? FYI, this happens occasionally, but by no means always.
Anyway, it is an interesting conundrum.
Mark
--miw
Without knowing how the
)
Without knowing how the validator exactly works, here is my guess.
If you check the results, especially the stderr, you will notice
that the checksums of the result files
Fstats.Ha: bytecount 1590349 checksum 76133021
Fstats.Hb: bytecount 1828464 checksum 87499938
differ from yours.
Fstats.Ha: bytecount 1590349 checksum 76293747
Fstats.Hb: bytecount 1828464 checksum 87685636
But this might be entirely wrong, so maybe Bruce can enlighten us.
And no this does not seem to be a Linux issue, because I use Linux too and
had, until now, no such problem.
HTH
Michael
Team Linux Users Everywhere
You have different
)
You have different checksums(look in your Result ID)than the other two who got credit...
Why?... No idea... :)
> So how about this one:
>
> http://einsteinathome.org/workunit/347835
>
> I returned the result in plenty of time. In fact, before one of the others
> returned their result, but where they were granted credit, my result was
> granted 0.0.
>
> Browsing through the results, it seem this is not altogether uncommon.
>
> I am assuming it is some kind of validation failure. The other two agreed and
> mine did not agrree with them. Could it have something to do with the fact
> that my result was computed under Linux and the others were computed under
> Windows? FYI, this happens occasionally, but by no means always.
>
> Anyway, it is an interesting conundrum.
>
> Mark
>