credit assignment problems

Message boards : Number crunching : credit assignment problems
Message board moderation

To post messages, you must log in.

AuthorMessage
Gorbag
Avatar

Send message
Joined: 1 Jul 99
Posts: 9
Credit: 18,763,310
RAC: 0
United States
Message 365 - Posted: 14 Jun 2004, 22:32:21 UTC

There still seem to be credit assignment problems. Check out this host:

http://setiboinc.ssl.berkeley.edu/sah/show_host_detail.php?hostid=731

ID: 365 · Report as offensive
SirUlli
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 58
Credit: 28,048
RAC: 0
Germany
Message 366 - Posted: 14 Jun 2004, 22:45:03 UTC

it is still Beta, and yea the Dev Team have to work about this and others, sometimes you got Credit after 2 succesfull Results, sometimes after 3 ...

Greetings from Germany NRW
Ulli
"I am convinced that human flight is possible and practical."
-- Wilbur Wright, 1899
ID: 366 · Report as offensive
Profile Darrell
Volunteer tester
Avatar

Send message
Joined: 14 Mar 03
Posts: 267
Credit: 1,418,681
RAC: 0
United States
Message 370 - Posted: 14 Jun 2004, 23:49:37 UTC

There is no validator problem. The first 9,000 units the project was populated with had the validator quorum only set to two units returned. Now that we are thru those the quorum has been set to three valid successful returns needed to get the unit validated. The problem is with the host that the link points to. Its benchmarks are 10 times slower than my 1ghz AMD Duron Windows 98 system.
ID: 370 · Report as offensive
Chuck

Send message
Joined: 3 Apr 99
Posts: 1
Credit: 0
RAC: 0
United States
Message 431 - Posted: 17 Jun 2004, 10:43:42 UTC - in response to Message 370.  
Last modified: 17 Jun 2004, 10:48:55 UTC

> There is no validator problem. The first 9,000 units the project was populated
> with had the validator quorum only set to two units returned. Now that we are
> thru those the quorum has been set to three valid successful returns needed to
> get the unit validated. The problem is with the host that the link points to.
> Its benchmarks are 10 times slower than my 1ghz AMD Duron Windows 98 system.
>
I tend to disagree. There is either a validator problem here too or all three hosts returned diferent results... all credit requests were reasonable close... no credit granted... my WU ID = 2958. It should have validated on the first two if it was a quorum setting.

How can results be examined at finer detail, or should this simply be passed forward to the Admin folks as a problem?

Here are more details to track with...

Result ID 8881
Name 11se03aa.14056.640.272152.141_2
Workunit 2958
Created 2 Jun 2004 20:55:41 UTC
Sent 12 Jun 2004 15:54:21 UTC
Received 13 Jun 2004 17:47:00 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x0)
Host ID 2872


Chuck

ID: 431 · Report as offensive
Profile Keck_Komputers
Volunteer tester
Avatar

Send message
Joined: 4 Jul 99
Posts: 1575
Credit: 4,152,111
RAC: 1
United States
Message 434 - Posted: 17 Jun 2004, 14:09:08 UTC - in response to Message 431.  

> > There is no validator problem. The first 9,000 units the project was
> populated
> > with had the validator quorum only set to two units returned. Now that we
> are
> > thru those the quorum has been set to three valid successful returns
> needed to
> > get the unit validated. The problem is with the host that the link points
> to.
> > Its benchmarks are 10 times slower than my 1ghz AMD Duron Windows 98
> system.
> >
> I tend to disagree. There is either a validator problem here too or all
> three hosts returned diferent results... all credit requests were reasonable
> close... no credit granted... my WU ID = 2958. It should have validated on
> the first two if it was a quorum setting.
>
> How can results be examined at finer detail, or should this simply be passed
> forward to the Admin folks as a problem?
>
> Here are more details to track with...
>
> Result ID 8881
> Name 11se03aa.14056.640.272152.141_2
> Workunit 2958
> Created 2 Jun 2004 20:55:41 UTC
> Sent 12 Jun 2004 15:54:21 UTC
> Received 13 Jun 2004 17:47:00 UTC
> Server state Over
> Outcome Success
> Client state Done
> Exit status 0 (0x0)
> Host ID 2872
>
>
> Chuck
>
>
The validate state of all three of those results is still initial, so either the validator has not run against them or they returned different results. The validator does not check based on requested credit, but one what the data returned is.



John Keck
testing BOINC since 2002/12/08
ID: 434 · Report as offensive
Profile Bob_Schonle

Send message
Joined: 17 Oct 01
Posts: 2
Credit: 656,286
RAC: 0
United States
Message 477 - Posted: 19 Jun 2004, 17:02:37 UTC

I find many of my work units that have finished have recieved zero credits, yet are considered done, over finished .
I am pretty discoraged that no credit is given, and I am damn sick and tired of the way this "Beta" system is working. If you programmers want people to join and do your work for free, you better come up with a decent scoring system, so we have something to feel good about, otherwise, you are a bunch of users and give little or nothing in return for our time, money and wear and tear on our equipment.

Bob Schonle
ID: 477 · Report as offensive
Heffed
Volunteer tester

Send message
Joined: 19 Mar 02
Posts: 1856
Credit: 40,736
RAC: 0
United States
Message 482 - Posted: 19 Jun 2004, 19:22:02 UTC - in response to Message 477.  

> I find many of my work units that have finished have recieved zero credits,
> yet are considered done, over finished .
> I am pretty discoraged that no credit is given, and I am damn sick and tired
> of the way this "Beta" system is working. If you programmers want people to
> join and do your work for free, you better come up with a decent scoring
> system, so we have something to feel good about, otherwise, you are a bunch of
> users and give little or nothing in return for our time, money and wear and
> tear on our equipment.

The concept of the beta test isn't about the credit. It's to find bugs. The credit system has some bugs.

As has been suggested elsewhere, if you are interested in credits, you should go run S@H1 instead of beta for BOINC.

And are you aware your credits will be wiped when BOINC is released?
ID: 482 · Report as offensive
Janus
Volunteer developer

Send message
Joined: 4 Dec 01
Posts: 376
Credit: 967,976
RAC: 0
Denmark
Message 483 - Posted: 19 Jun 2004, 19:28:25 UTC - in response to Message 477.  

> I find many of my work units that have finished have recieved zero credits,
> yet are considered done, over finished .

Well, then - name a few. My first look at your hosts didn't reveal any problems?
Except one of your hosts has benchmarked very low.

You are familiar with the fact that work will have to be checked before credit is granted right? - That's why it says 0 credit granted in some of the lists at the moment. You can see how much credit you have waiting by going to your userpage and pressing the "Pending credit" link.
ID: 483 · Report as offensive
Profile Bruno Moretti IK2WQA
Avatar

Send message
Joined: 15 May 99
Posts: 284
Credit: 49,167
RAC: 0
Italy
Message 485 - Posted: 19 Jun 2004, 21:37:57 UTC
Last modified: 20 Jun 2004, 2:27:31 UTC

My Total credit: 672,83
My Pending credit: 1128,47

Pending credit is pending because others users
have not returned their job.
Example

Meditation, not for Beta, but for the future:

IMHO if the work unit are assigned to three users only
and a user doesn't return its result,
nobody will ever get the credit for this wu.


Comments?


Clear skies from Italy!
ID: 485 · Report as offensive
Ingleside
Volunteer developer

Send message
Joined: 4 Feb 03
Posts: 1546
Credit: 15,832,022
RAC: 13
Norway
Message 487 - Posted: 19 Jun 2004, 22:16:40 UTC - in response to Message 485.  

> Meditation, not for Beta, but for the future:
>
> IMHO if the work unit are assigned to three users only
> and a user doesn't return its result,
> nobody will ever get the credit for this wu.
>
>
> Comments?
>

Then hitting the deadline, if only 2 of 3 is returned and 3 needed for credit, 1 new result is made and delivered to someone else, so this shouldn't really be a problem.


But looking on Chuck's wu 2958 here it's 3 returns but validator is standing on "initial". I really doubt the validator haven't run the last week, so something is wrong here. If only 2 of them or worse none was comparable, 1 or 2 more results should have been made for this wu.
Of course there's limits on how many total results for a wu, how many errored-out & how many good but no validation. But if someone haven't made a mistake and set this limit to 3 there is something wrong in one of the back-end-programs...

ID: 487 · Report as offensive
Profile Rom Walton (BOINC)
Volunteer tester
Avatar

Send message
Joined: 28 Apr 00
Posts: 579
Credit: 130,733
RAC: 0
United States
Message 489 - Posted: 20 Jun 2004, 5:43:25 UTC

This is another good reason to reset the database.

For the first couple of days this project was up, the file upload handler wasn't able to accept incoming result files. So basically the validator cannot validate the work-unit because one of the result files are missing.

ID: 489 · Report as offensive
Profile Bruno Moretti IK2WQA
Avatar

Send message
Joined: 15 May 99
Posts: 284
Credit: 49,167
RAC: 0
Italy
Message 501 - Posted: 20 Jun 2004, 18:01:25 UTC - in response to Message 489.  
Last modified: 20 Jun 2004, 18:15:17 UTC

> This is another good reason to reset the database.

Yes! I agree.

Another small problem for you, Rom
(and big problem for users):
upload is very, very fearfully slow.
Here and in BOINC Beta also.

Clear skies from Italy!
ID: 501 · Report as offensive
SURVEYOR
Volunteer tester

Send message
Joined: 19 Oct 02
Posts: 375
Credit: 608,422
RAC: 0
United States
Message 502 - Posted: 20 Jun 2004, 18:34:07 UTC

"Another small problem for you, Rom
(and big problem for users):
upload is very, very fearfully slow.
Here and in BOINC Beta also."

And In Boinc Alpha
Alpha Tester
Beta Tester
ID: 502 · Report as offensive
Profile KWSN - MajorKong
Volunteer tester
Avatar

Send message
Joined: 5 Jan 00
Posts: 2892
Credit: 1,499,890
RAC: 0
United States
Message 523 - Posted: 21 Jun 2004, 12:36:50 UTC

I wonder if the upload speed problem we are witnessing here (and in the Beta and Alpha projects) has any relation to the flakey router problem that S@H-Classic has at the moment between their ISP and the SSL building? That bad router has been hoseing connections over there for DAYS, and the natives over there are restless. The project staff have called in a repair order, the the ETA on the repair is sometime tomorrow. Just a thought.
ID: 523 · Report as offensive
Profile Guido_Waldenmeier_BiV

Send message
Joined: 3 Apr 99
Posts: 37
Credit: 0
RAC: 0
Germany
Message 524 - Posted: 21 Jun 2004, 13:56:19 UTC

allways people write BETA Project but Credits System are a Early ALPHA Thing mabye in 1 Year will work fine ,see SETI 1 in 1999- the same people the same story- ! i think
[/url]

Entwicklung von Verpackungen aus Wellpappe
ID: 524 · Report as offensive
Profile Bruno Moretti IK2WQA
Avatar

Send message
Joined: 15 May 99
Posts: 284
Credit: 49,167
RAC: 0
Italy
Message 907 - Posted: 24 Jun 2004, 0:53:54 UTC - in response to Message 485.  

> IMHO if the work unit are assigned to three users only
> and a user doesn't return its result,
> nobody will ever get the credit for this wu.

IMHO this problem is not resolved.
ID: 907 · Report as offensive
Profile mikey
Volunteer tester
Avatar

Send message
Joined: 17 Dec 99
Posts: 4215
Credit: 3,474,603
RAC: 0
United States
Message 965 - Posted: 24 Jun 2004, 3:11:23 UTC - in response to Message 907.  

> > IMHO if the work unit are assigned to three users only
> > and a user doesn't return its result,
> > nobody will ever get the credit for this wu.
>
> IMHO this problem is not resolved.
If the results are not received by the "report deadline" that is listed under the "work" tab in the Boinc program, then the Boinc Schedular resends those out to other computers automatically. Of course the Schedular KNOWS when it sent the units out and when they are due back. We went thru this in the Beta part of the program and it all worked out.


ID: 965 · Report as offensive

Message boards : Number crunching : credit assignment problems


 
©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.