[Info-vax] Raxco VMS Tuning Seminar Notes

[AEF]

On Nov 10, 2:22 pm, jls <notva... at yahoo.com> wrote:
> On Mon, 8 Nov 2010 16:49:22 -0800 (PST), AEF <spamsink2... at yahoo.com>
> wrote:
>
>
>
> >On Nov 8, 3:06 pm, jls <notva... at yahoo.com> wrote:
> >> On 3 Nov 2010 16:33:45 -0600, koeh... at eisner.nospam.encompasserve.org
>
> >> (Bob Koehler) wrote:
> >> >In article <gr73d61ue65tf9uiv65fhdu6mg8ihqa... at 4ax.com>, jls <notva... at yahoo.com> writes:
>
> >> >> PFRATL does not trim idle processes.  It is a quantum-driven memory
> >> >> management technique and idle processes do not reach quantum end.
>
> >> >   So they eithre got trimmed before they went idle, or when my process
> >> >   watcher called $GETJPI, which queues an AST to them.
>
> >> If they get paged down to swpoutpgcnt, then when they become active
> >> again, they will have to page in each page separately, and if these
> >> are hard faults, it will be one I/O for each page.
>
> >I recall that the manuals say that swapping is a very expensive
> >operation I/O wise. Am I missing something here?
>
> Old info - probably pre-V4.x.
>
>
>
> >That's why there's
>
> >3.7 Active Memory Reclamation from Idle Processes
>
> Which is fine as far as it goes, but if you page out lots of process
> pages, eventually the process will become active and those pages will
> be hardfaulted back in, one at a time.

What in particular are you referring to with paging out logs of
process pages? Why can't some of these pages come from the modified
page list?

>
> Again I reiterate - the swapper will bring in the entire process in
> ONE I/O.

Please explain "ONE I/O". (It's been a while since I've been up on all
this. One I/O for an entire process? You still have a certain amount
data go read in. And do you mean to imply that there is "one I/O" for
each page when they are faulted in?

I always inherited systems that were on their way out, and memory
wasn't really a problem. There were other problems of much greater
import.

Thanks for your continued comments.

AEF