From owner-freebsd-current@FreeBSD.ORG Tue Apr 1 18:03:04 2008 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 1CEB3106566B; Tue, 1 Apr 2008 18:03:04 +0000 (UTC) (envelope-from peter.schuller@infidyne.com) Received: from smtp.infidyne.com (ds9.infidyne.com [88.80.6.206]) by mx1.freebsd.org (Postfix) with ESMTP id C97578FC16; Tue, 1 Apr 2008 18:03:03 +0000 (UTC) (envelope-from peter.schuller@infidyne.com) Received: from c-a916e555.03-51-73746f3.cust.bredbandsbolaget.se (c-a916e555.03-51-73746f3.cust.bredbandsbolaget.se [85.229.22.169]) by smtp.infidyne.com (Postfix) with ESMTP id 0473D7DA10; Tue, 1 Apr 2008 20:03:02 +0200 (CEST) From: Peter Schuller To: freebsd-current@freebsd.org, d@delphij.net Date: Tue, 1 Apr 2008 19:05:21 +0100 User-Agent: KMail/1.9.7 References: <47F133AE.2010900@delphij.net> In-Reply-To: <47F133AE.2010900@delphij.net> MIME-Version: 1.0 Content-Type: multipart/signed; boundary="nextPart1774646.OXeITV98u8"; protocol="application/pgp-signature"; micalg=pgp-sha1 Content-Transfer-Encoding: 7bit Message-Id: <200804012005.32581.peter.schuller@infidyne.com> Cc: Ivan Voras Subject: Re: Are large RAID stripe sizes useful with FreeBSD? X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 01 Apr 2008 18:03:04 -0000 --nextPart1774646.OXeITV98u8 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline > Short answer: yes. Larger stripe sizes would allow better performance > for continuous read/write. In theory they will negatively affect small > size read/write but in practice you will notice that larger stripe size > would give better performance in almost all cases. One case where I have used large stripe sizes (though not with hardware rai= d=20 and not with FreeBSD) is with concurrent small reads (small =3D seek is the= =20 dominant cost) on top of a device of many drives. In these cases you want t= o=20 minimize the chance that a particular read will span multiple stripes, such= =20 that only a single drive will take a seek hit for a given request. This in= =20 turn maximizes actual physical concurrency when under high load (=3Dhigh=20 incoming request concurrency). =2D-=20 / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller ' Key retrieval: Send an E-Mail to getpgpkey@scode.org E-Mail: peter.schuller@infidyne.com Web: http://www.scode.org --nextPart1774646.OXeITV98u8 Content-Type: application/pgp-signature; name=signature.asc Content-Description: This is a digitally signed message part. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.8 (FreeBSD) iEYEABECAAYFAkfyeWwACgkQDNor2+l1i32B8ACgzhCc06rY/6YTWkPPOB/kVBQk Q3MAn2FqwhLtk3RoF9dhTjQ69GVf+Nw8 =a3dk -----END PGP SIGNATURE----- --nextPart1774646.OXeITV98u8--