From owner-freebsd-arch Sun Oct 14 21:10:21 2001 Delivered-To: freebsd-arch@freebsd.org Received: from peter3.wemm.org (c1315225-a.plstn1.sfba.home.com [24.14.150.180]) by hub.freebsd.org (Postfix) with ESMTP id 8B70537B409; Sun, 14 Oct 2001 21:10:09 -0700 (PDT) Received: from overcee.netplex.com.au (overcee.wemm.org [10.0.0.3]) by peter3.wemm.org (8.11.0/8.11.0) with ESMTP id f9F4A9M98707; Sun, 14 Oct 2001 21:10:09 -0700 (PDT) (envelope-from peter@wemm.org) Received: from wemm.org (localhost [127.0.0.1]) by overcee.netplex.com.au (Postfix) with ESMTP id 12A55380F; Sun, 14 Oct 2001 21:10:09 -0700 (PDT) (envelope-from peter@wemm.org) X-Mailer: exmh version 2.3.1 01/18/2001 with nmh-1.0.4 To: Matt Dillon Cc: "David O'Brien" , cvs-committers@FreeBSD.ORG, freebsd-arch@FreeBSD.ORG Subject: Re: cvs commit: src/sbin/newfs newfs.8 newfs.c In-Reply-To: <200110150256.f9F2ur151690@earth.backplane.com> Date: Sun, 14 Oct 2001 21:10:05 -0700 From: Peter Wemm Message-Id: <20011015041009.12A55380F@overcee.netplex.com.au> Sender: owner-freebsd-arch@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Matt Dillon wrote: > > :On Sun, Oct 14, 2001 at 04:03:03PM +0200, Ollivier Robert wrote: > :> Speaking of changing that value, shouldn't we also change the default bloc k > :> and fragment values ? > :> > :> It has been said in the lists many times that 16k/2k is more efficient > :> (and I'm using it myself by defautl now). > : > :There has been rummored problems if you use something other than 8k/1k. > :This is probably something that should run thru -arch or -hackers before > :doing it. > :"-c" was a no-brainer as noone has ever argued that a low "-c" was > :prefered (that I've seen). > > You can use 16K/2K safely. Anything larger may fragment the buffer > cache's KVA space and create issues. There are no known bugs (other > then fragmentation), but people have sporatically reported weirdness > with other combinations. There have been no solid bug reports. A ratio of 8:1 is the key for some reason. 4:1 used to regularly corrupt news filesystems a couple of years ago. (we used 4k/1k). Personally, I'm worried about using 16k/2k on anything less than a large (say larger than 1G) file system. If we made the defaults adjust to the fs size, I think that would be nice. (ie: default to max -c possible, and switch to 16k/2k for "big" fs's) Cheers, -Peter -- Peter Wemm - peter@FreeBSD.org; peter@yahoo-inc.com; peter@netplex.com.au "All of this is for nothing if we don't go to the stars" - JMS/B5 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message