From owner-freebsd-stable@FreeBSD.ORG Tue Apr 5 14:37:31 2011 Return-Path: Delivered-To: freebsd-stable@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 768CF1065676 for ; Tue, 5 Apr 2011 14:37:31 +0000 (UTC) (envelope-from erikt@midgard.homeip.net) Received: from ch-smtp05.sth.basefarm.net (ch-smtp05.sth.basefarm.net [80.76.153.6]) by mx1.freebsd.org (Postfix) with ESMTP id 07D298FC21 for ; Tue, 5 Apr 2011 14:37:30 +0000 (UTC) Received: from c83-255-51-20.bredband.comhem.se ([83.255.51.20]:41350 helo=falcon.midgard.homeip.net) by ch-smtp05.sth.basefarm.net with esmtp (Exim 4.73) (envelope-from ) id 1Q77AV-0004c4-IL for freebsd-stable@FreeBSD.org; Tue, 05 Apr 2011 16:23:48 +0200 Received: (qmail 67302 invoked from network); 5 Apr 2011 16:23:41 +0200 Received: from owl.midgard.homeip.net (10.1.5.7) by falcon.midgard.homeip.net with ESMTP; 5 Apr 2011 16:23:41 +0200 Received: (qmail 91723 invoked by uid 1001); 5 Apr 2011 16:23:41 +0200 Date: Tue, 5 Apr 2011 16:23:41 +0200 From: Erik Trulsson To: Pete French Message-ID: <20110405142341.GA91693@owl.midgard.homeip.net> References: <4D9B1E50.9020403@FreeBSD.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) X-Originating-IP: 83.255.51.20 X-Scan-Result: No virus found in message 1Q77AV-0004c4-IL. X-Scan-Signature: ch-smtp05.sth.basefarm.net 1Q77AV-0004c4-IL 8a8bd891cb07046cbed37a5e3b3ace74 Cc: spawk@acm.poly.edu, freebsd-stable@FreeBSD.org, freebsd@jdc.parodius.com, avg@FreeBSD.org Subject: Re: Kernel memory leak in 8.2-PRERELEASE? X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 05 Apr 2011 14:37:31 -0000 On Tue, Apr 05, 2011 at 03:04:22PM +0100, Pete French wrote: > > Adding some swap would help a lot more. > > So, I run a lot of systems without swap - basically my > thinking at the time I set them up went like this. > > "I have 4 gig of memory, and 4 gig of swap. Surely running 8 gig of > memory and no swap will be just as good ?" > > but, is that actually true ? Is real RAM as good as an equivalent amount > of swap, or is there smething special about swap which means you shoud > have some no matter how much RAM you have ? I believe some things (caches/buffers and the like) are sized according to how much real RAM you have, i.e. if you have 8G RAM the system will actuallu use more memory than if you have only 4G RAM. I also think that parts of the system are designed with the assumption that there is some swap available that can act as some sort of "overflow buffer" from time to time. -- Erik Trulsson ertr1013@student.uu.se