I see a big difference (3-4x) between Recent Average Credit and the average of changes in total credit for ClimateChange runs. I don't see similar differences for SETI.

I'm running CC on 3 computers (2 Macs, 1 Windows). On the computers, BOINCManager reports a host average of around 4,800 total (computers around 1800, 1650, and 1350). This agrees with the RAC reported on BOINCStats, which was 4,858 this morning. These numbers have been pretty stable for the past 10 days or so, going down a bit the day after we had power problems for several hours.

When I add up the total credit, and look at how that changes from day to day, I see only about 1,500 per day (computers around 650, 550, and 300). Both BOINCManager and BOINCStats agree on this number, as does boincsynergy when looking at just CC stats.

These two numbers differ by more than a factor of three, and that factor has been stable for the same time.

I have another computer that runs SETI. On it, the RAC, from several sources, shows up as 75-80, and the average of the changes in the total credit as 80-90.

What can I look at to help me understand why CC shows such differing results, or to diagnose whatever problems I might be having?

Thanks,

Pat

---

The computer with the rac that doesn't make sense is probably running HADAM3P models.
The problem was discussed here for a while.

What can I look at to help me understand why CC shows such differing results, or to diagnose whatever problems I might be having?

Looking at your computers, you have been working mainly on the 'HADAM3P' range of models.

We noticed a month or two back that these models seem to add a disproportionate amount to users' RAC, and discussed it at some length in the thread HADAM3P's too much RAC weight?. If you have any further idea what might have gone wrong, feel free to join the discussion ;-)

But seriously - whathever problem it is, it's a problem at the project's end not a problem with your computers - you don't have to worry about it.

Les and Richard,

Thanks for the explanation. As for the resolution, I'll just wait and see what happens. I'm assuming that the total credit numbers are correct and that eventually the RAC numbers will converge to them.

Pat

---

Yes, we are receiving the correct credits for every model type.

---

Cpdn news

You might want to check this thread about running 3P models on the Mac.

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=6595