# From michaelv@MindBender.serv.net Sat Sep 14 15:59:25 1996
#
# Actually, it looks more to me like it isn't paging aggresively enough
# to begin with.  So, when a heavy demand for memory comes all of a
# sudden, it has to work too hard to try and free up enough memory.
# What is notable is that it usually takes several of these multi-second
# pauses for it to free up enough memory for everything to run again, in
# many cases.
# 
# This is just how it "feels" to me.  I don't personally know anything
# about the vm code.

To hazard more guesses at details, it gives the appearance that the
paging scan rate doesn't exist at all; if it does, it's totally inef-
fectual.  I would figure that pages should be swept once in a while,
freeing memory, even when there is very little consuming memory.

I agree with the concept that backing store is not necessarily the way
to go, but we NEED to page things out.  Isn't there a hiwat/lowat setting
for paging/swapping in that if you exceed (hiwat)% of physmem, the scanner
actually starts touching pages and then paging them out, and when it drops
below (lowat)%, the scanner no longer touches the pages?

This would be a boon.  I've currently only got a lowly SLC with
16MB in it, but as I don't use it all that intensely at the moment,
I'm almost never above 50% of physmem, in which case not swapping is
kind of nice, but as soon as I start hitting, say, 9.6MB of physmem usage
(or availmem), the pager should start _gently_ nudging things out of
the way.

# 
# -----------------------------------------------------------------------------
#   Michael L. VanLoon                           michaelv@MindBender.serv.net
#         --<  Free your mind and your machine -- NetBSD free un*x  >--
#     NetBSD working ports: 386+PC, Mac 68k, Amiga, Atari 68k, HP300, Sun3,
#         Sun4/4c/4m, DEC MIPS, DEC Alpha, PC532, VAX, MVME68k, arm32...
#     NetBSD ports in progress: PICA, others...
# -----------------------------------------------------------------------------
# 

				--* greywolf;