From owner-freebsd-current  Tue Oct 29 00:54:03 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id AAA08176
          for current-outgoing; Tue, 29 Oct 1996 00:54:03 -0800 (PST)
Received: from dyson.iquest.net ([198.70.144.127])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id AAA08170;
          Tue, 29 Oct 1996 00:54:00 -0800 (PST)
Received: (from root@localhost) by dyson.iquest.net (8.8.2/8.6.9) id DAA03148; Tue, 29 Oct 1996 03:53:51 -0500 (EST)
From: "John S. Dyson" <toor@dyson.iquest.net>
Message-Id: <199610290853.DAA03148@dyson.iquest.net>
Subject: Re: Request to add this to FAQ re: swap space
To: rgrimes@gndrsh.aac.dev.com (Rodney W. Grimes)
Date: Tue, 29 Oct 1996 03:53:51 -0500 (EST)
Cc: dyson@FreeBSD.org, current@FreeBSD.org
In-Reply-To: <199610290837.AAA01517@GndRsh.aac.dev.com> from "Rodney W. Grimes" at Oct 29, 96 00:37:29 am
Reply-To: dyson@FreeBSD.org
X-Mailer: ELM [version 2.4 PL24 ME8]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-current@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

> > > 
> > > > as usual, add some space in for expansion and for process growth.
> > > 
> > > Thats what ``swapon'' is for :-) :-) :-) :-)
> > > 
> > So are you going to add the code to properly support swapon for files? :-).
> 
> I thought this was long ago working, what happened?
> 
We had experimented with it, but I have been hoping to implement a proper
swapon/swapoff.  There are only so many hours in a day :-).

>
> > I think that 2.2 is going to be frozen soon, and the risk/reward might be
> > excessive to do it.  VN devices exact a significant performance hit.  In
> > fact, the swapon for files will also.
> 
> ``swaping/paging'' imposes enough of a performance hit that my methodology
> already takes care of that, just don't swap more than you have memory,
> I rarely wait for my swap space usage to hit 50% before I toss more memory
> in a box.  Of cource, the boarder cases already talked about are an exception
> to this condition (can't put enough physical RAM in most machines for them
> to be at 1 x RAM swapsize).
> 
The issue is that there are alot of individuals who run X in 16-32MB and
will complain if they cannot have 20-30 X applications running (perhaps
inactive, on swap space.)  You can easily run out of swap space in that
condition.  Those people who know that they will not run out of swap
space, can configure their systems appropriately.  Note that processes
that are idle and paged/swapped out take no system time.  It is mostly
when a system is thrashing that more memory is needed.

John