climateprediction.net home page
Scheduler request too recent

Scheduler request too recent

Message boards : Number crunching : Scheduler request too recent
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3

AuthorMessage
Jean-David Beyer

Send message
Joined: 5 Aug 04
Posts: 1045
Credit: 16,506,818
RAC: 5,001
Message 61946 - Posted: 10 Jan 2020, 3:17:33 UTC - in response to Message 61894.  

Which doesn't take into account that most people joining over the past couple of years seem to have no particular interest in the climate models, and load up with lots of work from other projects as well.


I joined quite a while ago. I am particularly interested in climate models so my priorities are 50% climateprediction, 25% WCG, 13% Rosetta, and 12% Seti@home. When I can get all the work units I wanted, this ended up with three processors running CPDN, and the others taking their turns on the other processor.

These days there are large gaps in work-unit availability, so now I have Rosetta and Seti@home set to Won't Get New Tasks. I allow new tasks in CPDN until I get four work units, and then turn off getting new ones. I leave WCG on, but I accept only one work unit at a time on two of their jobs, and two at a time on one other. As a result, I seldom get any.

Right now, my deficit on CPDN work is so great that I am usually getting four work units at a time. (Right now I am running one N144 work unit and three N216 work units.) But when this does not happen, I get a work unit or two from WCG that usually go through in less than a day. If I run out of work, I enable either seti@home or rosetta just long enough to get some work units (usually 6 or 8 come through) and they grab the processors. By the time they are done, I have received some more CPDN work. If work units were available more of the time, I hope to find a way to just let them all run all the time without my fiddling around like this. I.e., like in the old days.
ID: 61946 · Report as offensive     Reply Quote
Profile Dave Jackson
Volunteer moderator

Send message
Joined: 15 May 09
Posts: 4314
Credit: 16,380,160
RAC: 3,563
Message 61948 - Posted: 10 Jan 2020, 7:11:42 UTC

You need to set the resource share in the computing preferences for each project on your account page for that project.


If you set resource share to 0 for a project it the boinc server code should only send work if there is no work for your other project(s). Unfortunately I learned from the BOINC forums not all projects allow a 0 resource share and have their system set up to allow a minimum of 1.

https://boinc.berkeley.edu/forum_thread.php?id=13359
ID: 61948 · Report as offensive     Reply Quote
Previous · 1 · 2 · 3

Message boards : Number crunching : Scheduler request too recent

©2024 climateprediction.net