Message boards :
Number crunching :
BOINC 7.0 released to public
Message board moderation
Previous · 1 . . . 3 · 4 · 5 · 6
Author | Message |
---|---|
David S Send message Joined: 4 Oct 99 Posts: 18352 Credit: 27,761,924 RAC: 12 |
I am just curious, as I have NO intent of switching. My questions apply to both REC and the debt system, I guess. How does it apply in a situation where a project I'm attached to doesn't have any work for a long time? Like ever. I attached to Orbit sometime last year, but it has never had any work to send me. If it suddenly does generate some work, will my machines remember all that accumulated debt and run Orbit exclusively to make up for it? Also, does debt/REC work on each individual machine, or does some server somewhere monitor all your hosts and average it out? David Sitting on my butt while others boldly go, Waiting for a message from a small furry creature from Alpha Centauri. |
LadyL Send message Joined: 14 Sep 11 Posts: 1679 Credit: 5,230,097 RAC: 0 |
6.10.60 - been a loong time So I'm not confident on it. Debt was supposed to stop hosts fetching tasks if it got too large. But Richard Haselgrove has debt logging that shows it's not really working. IIRC it will still have one task per project in the queue. even if giving Orbit more room. if you were on 7.0.25 it would see that REC is 0. there for priority is 0 (for some reason priorities are negative i.e. 0 is highest prioroty) So the project would get asked first for work on every workfetch. If it got work, it would indeed devote most resources to catching up - but only up to an REC that coressponds to its resource share. As REC goes higher priority drops until the other project are in front again. I'm not the Pope. I don't speak Ex Cathedra! |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
But Richard Haselgrove has debt logging that shows it's not really working. Actually, I'm still running the logger on my v6.12.34 hosts - though I'll have folders full of unanalysed CSV files, dating back years. Every time I notice something not fetching, and look at a live snapshot, there's a reasonable explanation. And yes, I'm attached to orbit too (which never has work), and to LHCclassic, which puts through a few tasks once in a blue moon - all get processed OK. I set up the logging process because there was a big debt leak in the early CUDA-capable clients - around 6.6, probably. But David eventually got the hint and fixed it... |
TRuEQ & TuVaLu Send message Joined: 4 Oct 99 Posts: 505 Credit: 69,523,653 RAC: 10 |
You say on every workfetch...?? Isn't the "communication deferred" the prefferred thing to use? Like having 0 tasks of a high resource project waiting for the "communication deferred" to countdown to 0 and then request new tasks?+ Why should a project wait for "workfetch" to start requesting tasks.? What is the benefit of that(for the user that awaits "requesting new work"? //TQ TRuEQ & TuVaLu |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.