There's a user with a large number of Macs who hasn't been getting credits for weeks.
Is anyone else having problems lately with no credits?

No credit since May; 19 completed tasks in that time. Computer ID is 1233059.

---

Email sent.

I've just had a reply to the email that I sent:

The project people are working hard at trying to understand why the new Mac version behaves as it does, and then to fix it.

The script will be manually run until such time as a working version of the program is released.

all my macs got no credit or show trickle on webpage for wah2

completed tasks - no credit!
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19537644
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19603818
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19621758
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19680809
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19701200


task running some time - but not ready, no credit
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19400107
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19551323
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19635737
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19646010
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19654774
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19682279
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19720074
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19746212
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19790366

maybe they need to investigate.

- while I'm waiting and hoping ;)

---

I have the same problem that the Claimed Credit for all wah2 tasks shows as '---' and the Granted Credit is 0.00 - with approximately 60 wah2 tasks completed and several currently being processed.

From a quick read of some of the other posts, I'm unclear whether I need to ask that someone runs the appropriate script or whether my tasks will be spotted in an automatic sweep? If the latter, they haven't been so far.

It isn't a huge issue for me, but it does mess up the allocation of time between CPDN and other projects.

I'd be grateful if you could clarify the position.

Thanks.

---

Hi Brian

See my post, 4th from the start of this thread.

I've had a quick scan of people who have posted here, and there's credits for lots that have been completed since May.

The export files have also been updated in the general run for this weekend, so I'll look at BOINCstats in a day or 2.

My credit is up to date now, thanks! (The "recent average credit" didn't change, but the total is updated. Progress!)

---

Yes, all my wah2 wu's which successful ended, got credit!
many thanks.

Hopefully they find and repair the bug soon, preventing the mac app from trickling
– so no manually action is needed anymore.

---

Hi Les - I can confirm that all of the completed wah2 tasks now have credit attached to them - thank you to all involved.

One follow up observation. BOINC still appears to allocating almost all calculation capacity to this project - is that because it focuses on the average work done rather than any other measure?

Thanks again.

---

Hi Brian

It's because of the very long run times, compared to the quite short ones of other projects.

At the start of each new climate model type, BOINC doesn't have much of an idea of how long it will take to run them, so when there's a mix of tasks from various projects, BOINC plays it save, by running the climate model most of the time, while it "learns" this.
Once it finds out, it will start to give more time to other projects.

This only works if no changes are made to BOINC options, as then it will have to start the learning process again. Although it may not take as long this time.

Included with each task is an estimate of how long it will take.
But this can be very wrong, and it's for a certain speed of computer, with a certain mix of tasks.

In the case of climate models, back when we still had a beta site, several of us were providing information about this to try and get it right.
For a certain computer speed, (usually "fast"), running a certain number of models, and, usually, no work from other projects.

the average credit does not move since May

the average credit does not move since May

The credits for the current Mac version of the WAH2 application are not based on trickles because a bug in that application prevents the creation of trickles. The final credits are awarded by some other process but I assume that the average credit calculation still depends on trickles - so a Mac running WAH2 will find its average credit reducing to zero even though completed models do receive the right credits. That's what's happening to my Mac.

I wonder what happens if the run doesn't complete ...

If a wu doesn't finish because of an error it doesn't get any credit :(

- as I can see for some of my wu's:
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19599378
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19649365
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19722630

I hope there will be some process to get credit for wu's which did running some time even if they stop before successful end.

---

If a wu doesn't finish because of an error it doesn't get any credit :(

- as I can see for some of my wu's:
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19599378
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19649365
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=19722630

I hope there will be some process to get credit for wu's which did running some time even if they stop before successful end.

Thanks for that. Pretty difficult to count Zip files when they don't necessarily go to the same place as the trickles or the restart Ziip (or even the same continent). More reason to fix the trickle bug, which is being addressed.

I've just had a reply to the email that I sent:

The project people are working hard at trying to understand why the new Mac version behaves as it does, and then to fix it.

The script will be manually run until such time as a working version of the program is released.

Are there any news about the Mac version?

Btw: I found some wah2-wu get about double the credit as other:
ID: 19832510 cpu-time: 231,012.97 Credit: 9,900.12
ID: 19830715 cpu-time: 231,486.58 Credit: 9,900.12
ID: 19746212 cpu-time: 597,006.82 Credit: 9,900.12
ID: 19646010 cpu-time: 575,867.35 Credit: 9,900.12

ID: 19701200 cpu-time: 381,513.65 Credit: 18,277.15
ID: 19680809 cpu-time: 396,254.01 Credit: 18,277.15
ID: 19678778 cpu-time: 303,213.78 Credit: 18,277.15

looks really strange - as some did run about double the time but did get only about half the credit!
May be someone could explain me.

---

Btw: I found some wah2-wu get about double the credit as other:
ID: 19832510 cpu-time: 231,012.97 Credit: 9,900.12
ID: 19830715 cpu-time: 231,486.58 Credit: 9,900.12
ID: 19746212 cpu-time: 597,006.82 Credit: 9,900.12
ID: 19646010 cpu-time: 575,867.35 Credit: 9,900.12

ID: 19701200 cpu-time: 381,513.65 Credit: 18,277.15
ID: 19680809 cpu-time: 396,254.01 Credit: 18,277.15
ID: 19678778 cpu-time: 303,213.78 Credit: 18,277.15

looks really strange - as some did run about double the time but did get only about half the credit!
May be someone could explain me.

The second group of models in your list run for 24 model months and (would) therefore generate 24 trickles, whereas the other models run for 13 months and (would) therefore generate 13 trickles. Since credits are awarded for trickles, models in the second group get twice as many credits.

The CPU time for each model type depends on a number of factors. As well as the number of simulated months, the model grids are different sizes (i.e. the region covered (Europe, Mexico etc.) and the resolution [e.g. 25 km or 50 km]), the input data files may be more or less complicated, different variables may be calculated for reporting, and the sea/atmosphere may have a greater or fewer number of levels. So the trickle method of allocating credits won't work very well when one application behaves so differently. In the past there tended to be one application per experiment and the credit weighting could be varied per application, but with the new-generation generic applications (e.g. WAH2) that flexibility is removed.

Thank you for the explanation.

lately I didn't follow the project news because lack of time.
I saw different Credit/CPU-time in the past but couldn't remember that it was so much.

---

I am having the same issue. My iMac runs but doesn't add anything to the 'Recent Credit" value.