Hi,

I've been working on getting the crediting to stabilize. During these changes there was a time where valid results were given a near zero credit (it was shown as zero). This affected 290 results for 54 users and 80 hosts.

I've just fixed credit for these anomalies by granting them a fixed amount. Also, the near zero credits shouldn't happen anymore since it seems to have been result of a --max_granted_result switch for the validator that I tried as a solution for the credits. I'm not sure if it's a bug in BOINC Server code or if it's really supposed to work that way..

Otherwise, the crediting seems to have somewhat stabilized but I think we are giving too much credit at the moment. I'm going to bring it down a bit.

There are still cases were huge credit is given for some results. If validator doesn't correct itself I need to change how we credit results and bypass all these scaling that goes on.

-w

Unit 128829 received 0 credits, yet validated.

Please look into this.

Unit 128829 received 0 credits, yet validated.

Please look into this.

Is this request being ignored? Why did I get 0? Will you issue credit for this?

no, this request was not missed.
We totally changed the credit system, like you probably noticed.
That task will probably receive 502*7=3514 credits eventually.
There's a huge list of things to fix and new feature on the roadmap

Thanks for your patience and keep crunching ;)

Unit 128829 received 0 credits, yet validated.

Is this request being ignored? Why did I get 0? Will you issue credit for this?

Sure, it was on my list as soon as I saw your post and I finally managed to fix validator for these results yesterday. So this (and some other valid but zero credited results) now has credit granted.

I'm sorry that this took this long. Thanks for crunching! And for pointing me to this problem. :)

-w

Hello.Is that credit for a runtime less then a half hour normal?
My cards are 2 5850 on win 7 64bit.
Here an example:
dnetc_r72_1306251906_4_247
Anwendung Distributed.net Client
erstellt 24 May 2011 | 15:45:06 UTC
autorisiertes Ergebnis 319482
gewährte Punkte 2,099.00
Mindestanzahl 1
Initialreplikation 1
max # von Fehler/Gesamt/Erfolg Aufgaben 5, 10, 5
Greetings.

Hello.Is that credit for a runtime less then a half hour normal? ....

The credits are spot on, the rate is 8.5 per Stat unit. In your example, there are 247 stat units (look at the last three figures in the WU name) x 8.5 ...

The times do look a little on the slow side for 2x5850s, but that will depend on what you are running the cards at. Have a look at my second machine "Dot-PC", it has a single 5850 in there running at 775/500.

Individual preference - of course - on the running speed, and if you are running default settings that would probably go a long way to explain speed difference, but purely from credit angle, you have the correct credits being given (you can check back on ones completed - just multiply the last three figures by 8.5)

Its the same rate as given by DNETC, and personally I find this one more stable and not suffering from the long standing dual card problem of DNETC fame. I did a 24hr session recently over there as part of my "where do I stay" decision, and as far as I am concerned, for me, this app is more stable - everyones mileage will differ of course, but this app gets my Vote.

Regards
Zy

Its the same rate as given by DNETC, and personally I find this one more stable and not suffering from the long standing dual card problem of DNETC fame. I did a 24hr session recently over there as part of my "where do I stay" decision, and as far as I am concerned, for me, this app is more stable - everyones mileage will differ of course, but this app gets my Vote. Zy

Ditto, Moo! works better here too. Multiple cards work sub-optimally at both, but it seems that's a problem with the dnet client. Also a big difference for me is that we get straight answers here and never did at the other project.

[quote]Multiple cards work sub-optimally at both, but it seems that's a problem with the dnet client. /quote]

right!

you would need WU's with a single packet AND boinc running a seperate WU on each GPU.


not possible now imho..

....... you would need WU's with a single packet AND boinc running a seperate WU on each GPU...... not possible now imho..

Impossible to run singles on each GPU, and likely to remain so for foreseeable future.

The Core App is from Distributed.net, and it will take whatever GPUs are in the machine as a multi thread, it cannot be changed at this end by Moo! All that happens here is the BOINC Wrapper Application is put round it to make it distributable to BOINC outlets - and so far in my view the use of the BOINC Wrapper here has resulted in a more stable app, so thumbs up to Teemu :)

Regards
Zy

Having BIG problems with running crossfired 5870's here !
The box has 4x5870 (2 gig mem each). In 6 sec I have 47.6 % done, after that the GPU use stays at 90-95% for 30 minutes, then it starts bouncing between 0-25% for 26 minutes=runtime 56 minutes on 4x5870 @930 core & 1300 memory...with a nice 2159 in credits...I do a Collatz wu in 20 min/GPU with it, so 2159 in here and 36000 @ collatz in the same time...whats the problem with boxes with several crossfired GPUs ?

....... you would need WU's with a single packet AND boinc running a seperate WU on each GPU...... not possible now imho..

Impossible to run singles on each GPU, and likely to remain so for foreseeable future.

oh, really?

just not implemented.

the CPU app can run multiple instances and can be configured to use selected cores instead of running in MT-mode.

but that's a task for distributed.net...

Having BIG problems with running crossfired 5870's here !

Disable Crossfire. It made a big difference for my HD5830s. See here:

http://moowrap.net/forum_thread.php?id=41&nowrap=true#329