Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 07 Oct 2000 12:10:44 -0600
From:      Warner Losh <imp@village.org>
To:        Bill Moran <wmoran@columbus.rr.com>
Cc:        FreeBSD Stable <stable@FreeBSD.org>
Subject:   Re: `time make buildworld' 
Message-ID:  <200010071810.MAA01453@harmony.village.org>
In-Reply-To: Your message of "Sat, 07 Oct 2000 10:39:03 EDT." <39DF3587.3C2BA373@columbus.rr.com> 
References:  <39DF3587.3C2BA373@columbus.rr.com>  <39DECC00.71E81EF1@urx.com> <Pine.BSF.4.21.0010070011530.84656-100000@fremont.bolingbroke.com> <20001007101732.A91198@freebie.demon.nl> 

next in thread | previous in thread | raw e-mail | index | archive | help
In message <39DF3587.3C2BA373@columbus.rr.com> Bill Moran writes:
: Just to add to this, I've done some very limited experiments with
: softupdates and my conclusion is that running softupdates on systems
: with low RAM is a very bad idea (tm) as it hurts performance badly. On
: systems with a more reasonable amount of RAM it's an improvement.
: I say limited tests because I haven't run the same machine with
: low/sufficient RAM to verify that it's the RAM that causes the
: slowdown, but it makes sense.

I know that softupdates on my 16MB OpenBSD/arc system trippled the
make build time from 26 hours to 76.  When I put 16MB more into that
machine, for a total of 32MB, the build world time with stoft updates
and without droped to 21 and 19 respectively.  When I put another 32MB
into the machine (to max it out), the build world time dropped another
hour or so (17.5 with soft updates, 20.5 without).

On a 8MB system, I'd recommend against softupdates.  They need a lot
of RAM to be useful.

Warner



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200010071810.MAA01453>