climateprediction.net (CPDN) home page
Thread 'Why does BOINC client scheduler work differently for different projects?'

Thread 'Why does BOINC client scheduler work differently for different projects?'

Questions and Answers : Unix/Linux : Why does BOINC client scheduler work differently for different projects?
Message board moderation

To post messages, you must log in.

AuthorMessage
Jean-David Beyer

Send message
Joined: 5 Aug 04
Posts: 1120
Credit: 17,202,915
RAC: 2,154
Message 4159 - Posted: 15 Sep 2004, 16:12:30 UTC

I have two Hyperthreaded Xeon processors, so my RHEL 3 ES OS thinks it has 4 processors. I allow BOINC to use all 4. I have told BOINC to allow suspended applications to remain in memory (or to be paged out).

Now when climateprediction does that, what happens is that the two processes are put in STOPPED state and when the BOINC client sees fit to resume them, it does that. Thus, the times slowly increase.

But setiathome is different. It seems to put nothing in STOPPED state, yet it allows the processes to resume, accumulating time just as does climateprediction.

I do not know what Predictor does, since it will not run with the 4.05 BOINC client.

I would have though all applications would use the same mechanism for stop and resume. ???
ID: 4159 · Report as offensive     Reply Quote

Questions and Answers : Unix/Linux : Why does BOINC client scheduler work differently for different projects?

©2024 cpdn.org