this problem with this approach is that it unbalances the system
too much in favor of anonymous data.  if there is enough swap-backed
data to fill the 2/3 of memory which we attempt to keep in the active queue,
then vnodes will only have the 1/3 of memory in the inactive queue available
to cache their data, which is silly if the anonymous memory is (eg.)
netscape's in-memory cache and netscape has been iconified in the corner
of the screen all day while the user has been trying to do some file-
intensive work.

I do agree that the paging algorithm does need to be enhanced,
but this isn't the way to do it.    from subsequent mail on this list,
it appears that people are already figuring out that part of the problem
is a simple error on my part, but I haven't ready through all the rest
of this week's tech-kern mail yet, so I'm not sure where things stand
right now.

-Chuck



On Tue, Jan 23, 2001 at 07:34:10PM -0000, eeh@netbsd.org wrote:
> 
> Here's a little fix I threw together to improve the performance of
> UBC under I/O load.  What it does is put buffer cache pages on the
> uvm inactive list when an IOP completes instead of leaving them on
> the active list for the page scanner to reclaim.  This has several
> effects:  it makes buffer cache pages more likely to be reclaimed
> than other pages, and it slightly reduces the load on the pagedaemon.
> 
> Eduardo