one thing you could do is a quick crontab entry that suspends the big
mutha at 8am (or whatever) (send a kill -STOP pid to it) and then at
5pm (or whatever) send a kill -CONT to it. If you are memory bound
I don't see any other solution because when you 'go idle' in comes
the big job and out goes your desktop stuff...
Robert Hyatt Computer and Information Sciences
[EMAIL PROTECTED] University of Alabama at Birmingham
(205) 934-2213 115A Campbell Hall, UAB Station
(205) 934-5473 FAX Birmingham, AL 35294-1170
On Fri, 15 Jan 1999, Jeffrey Mark Siskind wrote:
> There was some recent discussin of `nice' and the desire to leave a
> long-running process at a high nice level and have it use spare cycles but not
> impede interactive use.
>
> I have use for this too. But with a twist. I don't really care whether the
> long-running job gets 0% of the CPU or 5%. But I do care about swapping. I
> have 1G on my machine. And my long-running process uses about 900M. And my
> interactive use typically uses 200M. So when my long-running process runs,
> it swaps out all of my desktop (emacs, etc). What I do now is every time I sit
> down at my machine I suspend the long-running process to get interactive
> response. And when I leave, I resume it. But sometimes I forget. And then I
> waste a whole night's worth of cycles. Ideally, I'd like to set something up
> that figures that if I didn't type at the console (or start a new process or
> something like that) for the past, say, five minutes, then declare the machine
> to be idle and OK to run my long-running-process. But immediately when I start
> to type (or whatever puts my machine back in `interactive' mode) page out my
> long-running-process, or at least don't let its page-in requests compete with
> my interactive process page-in requests.
>
> Basically, I think that the nice level needs to affect not only time slices in
> the scheduler but also page-in or disk-access priority. Is that possible?
>
> Jeff (http://www.neci.nj.nec.com/homepages/qobi)
> -
> Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
> To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
>
-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
