Conflict with SETI?
13 Dec 2014 22:40:52 UTC
Topic 197874
(moderation:
I have E@H work on my computer, some "ready to start" and some "waiting to run", but ever since SETI started issuing work units again, they seem to be the only ones that are being run. Time share is 50/50 and no suspend orders were given for E@H. Any ideas?
Conflict with SETI?
)
That is a BOINC function, neither SETI nor Einstein governing.
While the modern BOINC foreswears the old terminology of debt, it nevertheless has noticed that while you wanted both, you only got and ran Einstein for a while. So by whatever the modern replacement for debt is I think it is balancing things up a bit. I predict that if you leave it be, it will again run some Einstein--perhaps after a few weeks.
Thanks to a Richard
)
Thanks to a Richard Haselgrove reply in another thread, I can provide you a link to design documentation for the system which replaced debt.
credit-driven scheduling
RE: Thanks to a Richard
)
But do be aware that the measure used is Recent Estimated Credit (REC), not actual credit - so all the vagaries of pending tasks and credit rates are ignored.
You will find that over time the two projects (any two, or more, projects) will settle down and co-exist - but any perturbation, such as one project running out of work for a while, will upset the equilibrium for a while until REC equalises again.