From owner-freebsd-hackers Thu May 25 13:30:38 2000 Delivered-To: freebsd-hackers@freebsd.org Received: from fw.wintelcom.net (ns1.wintelcom.net [209.1.153.20]) by hub.freebsd.org (Postfix) with ESMTP id 82FA237BDCC for ; Thu, 25 May 2000 13:30:33 -0700 (PDT) (envelope-from bright@fw.wintelcom.net) Received: (from bright@localhost) by fw.wintelcom.net (8.10.0/8.10.0) id e4PL4k929591; Thu, 25 May 2000 14:04:46 -0700 (PDT) Date: Thu, 25 May 2000 14:04:46 -0700 From: Alfred Perlstein To: Matthew Dillon Cc: Anatoly Vorobey , hackers@FreeBSD.ORG Subject: Re: Proper uses for MFS? Message-ID: <20000525140446.J28594@fw.wintelcom.net> References: <200005251705.NAA67491@blackhelicopters.org> <200005251757.KAA83404@apollo.backplane.com> <20000525141623.D6776@sasami.jurai.net> <200005252022.NAA84015@apollo.backplane.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2i In-Reply-To: <200005252022.NAA84015@apollo.backplane.com>; from dillon@apollo.backplane.com on Thu, May 25, 2000 at 01:22:40PM -0700 Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG * Matthew Dillon [000525 13:58] wrote: > > :You, Matthew Dillon, were spotted writing this on Thu, May 25, 2000 at 10:57:33AM -0700: > :> > :> I don't particularly like to use MFS for 'large' partitions, mainly > :> because cached data blocks wind up in core memory twice (once in MFS's > :> memory map, and once in the VM page cache). > : > :You've said this several times in threads on MFS during recent months, > :and I've always wanted to ask: is that a necessary 'feature' of MFS's > :architecture, or something which could possibly be fixed without > :too much hard work? For instance, would it be possible to force > :VM not to cache MFS pages, etc.? > : > :-- > :Anatoly Vorobey, > > The double caching is a consequence of MFS's 'physical media' being > the mmap() rather then real physical media. It would be difficult > to fix, and probably not worth the effort. > > MFS's only advantage is that the double-caching tends to keep pages > in-core longer, and you have less 'real' physical I/O because things > like write-behind and buffer cache flushes are doing nothing more > then flusing from the kernel's main VM page cache into MFS's memory > map. > > If you have enough memory not to care about the double-caching issue, > then MFS will work fine. y'know... We could do a really good job of a disk backed MFS by having a mount flag for the syncer to ignore a mount point as well as marking it async. -Alfred To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message