I recently re-attached to Climateprediction.
I had problems in the past, But they seems to have disappeared. Which is good. I finished a job for over 700 hours without any problems.
However, the size of the project files are excessive. And it interferes with other projects. It prevents them from running.
I chose the option to not get any new tasks. But it ignored that. When the 700 hour task finished, a new task downloaded anyway. The size of the new one is 1058 hours. I have suspended Climateprediction because of this.
What happens to the downloaded file, if I do not run it?

---

the size of the project files are excessive

It\'s possible that you were last here before a lot of the new types of models (for different research), existed.
The size of the download (and upload) files is different for each of the model types, with those for the hadam3p being the biggest.
This is for the reasons given years ago when people were asking for smaller models: they are the longer models cut into shorter sections, and require a lot more data transfer in both directions to enable these short sections to be joined together.

It\'s possible to choose which models to run / (not to run), by clicking on a select box in the climateprediction.net prefs in your account.

**********************

And it interferes with other projects. It prevents them from running.

This is because of the long run times, compared to other projects.
It takes a while for BOINC to adjust itself to suddenly having to accommodate very long WUs with short WUs. At least several hours; possibly a few days.
And then it will start alternating the WUs from all projects, with a long term approximation of the resource shares that you\'ve set.

**********************

I chose the option to not get any new tasks. But it ignored that.

Odd. It\'s always obeyed that for others. I have it set that way for both of my computers and don\'t have problems.

**********************

What happens to the downloaded file, if I do not run it?

It will just sit there on your computer until it\'s either run or aborted.

Another thing that\'s changed about the project, is that models are no longer re-issued after a certain period without server contact.
A fixed number (which varies, depending on requirements), of each model are created at the start, with the hope that at least one will get run and completed.

---

Backups: Here