I've the problem that BOINC updates predictor@home WU's all the time, but there are no messages shown in boinc telling me, that Seti tried to get WU's.
If I delete predictor project, there are messages telling me, that there is no work and it will retry in $random_time.
Is that normal?

Do a thread search on "multiple projects"

This has been answered numerous times.

<a> [/url]

> Do a thread search on "multiple projects"
>
> This has been answered numerous times.

May be. But answers like these are completely useless. At least when given in a board named "getting started"...

As I have read e. g. in number crunching there are some problems with boinc working at multiple projects. I have not found a real answer but it seems that boinc will work only one project at a time which might leed to only work at predictor for the moment where seti will not provide any work. When getting to the low water mark it will try to get work from seti too. I may be wrong but this is what the comments sounded for me.

Bye
Peter

P.S. You might detach from predictor and then you will see boinc requesting work from seti (but not getting any as far as I can see). Then you can attach with seti again.

> May be. But answers like these are completely useless. At least when given in
> a board named "getting started"...
>
> As I have read e. g. in number crunching there are some problems with boinc
> working at multiple projects. I have not found a real answer but it seems that
> boinc will work only one project at a time which might leed to only work at
> predictor for the moment where seti will not provide any work. When getting to
> the low water mark it will try to get work from seti too. I may be wrong but
> this is what the comments sounded for me.

Sorry, caught me at a bad time. I personally have answered this question numerous times.

No, the problem isn't that there are problems with BOINC running multiple projects. It's just that people don't understand what is going on when they join a new project.

It is normal for the new project to dominate the WU queue until the resource share balances out between the projects.

It can also appear lopsided however, if the other project(s) don't currently have any work to give. In such a situation, the other project(s) will pick up the slack. In other words, If S@H can't give you anything, P@H will fill your entire queue. Then when the low watermark is hit and BOINC requests more work and S@H is still having issues, you will again get a full queue of P@H.

<a> [/url]