From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from firewall.xsintricity.com (ovpn01.gateway.prod.ext.phx2.redhat.com [10.5.9.1]) by int-mx04.intmail.prod.int.phx2.redhat.com (8.13.8/8.13.8) with ESMTP id o4RGex2B007746 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO) for ; Thu, 27 May 2010 12:41:20 -0400 Received: from firewall.xsintricity.com (firewall.xsintricity.com [172.31.0.1]) (authenticated bits=0) by firewall.xsintricity.com (8.14.3/8.14.3) with ESMTP id o4RGewPi005297 (version=TLSv1/SSLv3 cipher=DHE-RSA-CAMELLIA256-SHA bits=256 verify=NO) for ; Thu, 27 May 2010 12:40:58 -0400 Message-ID: <4BFEA099.9020005@redhat.com> Date: Thu, 27 May 2010 12:40:57 -0400 From: Doug Ledford MIME-Version: 1.0 References: <4BF5A883.7060503@tlinx.org> <20100521051021.GA1412@maude.comedia.it> <4BF62CBF.3070702@tlinx.org> <20100522072321.GB12294@maude.comedia.it> In-Reply-To: <20100522072321.GB12294@maude.comedia.it> Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="------------enig5A1AB7CA1D2F6322B4DE4D92" Subject: Re: [linux-lvm] Volume alignment over RAID Reply-To: LVM general discussion and development List-Id: LVM general discussion and development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , List-Id: To: linux-lvm@redhat.com This is an OpenPGP/MIME signed message (RFC 2440 and 3156) --------------enig5A1AB7CA1D2F6322B4DE4D92 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable On 05/22/2010 03:23 AM, Luca Berra wrote: > On Thu, May 20, 2010 at 11:48:31PM -0700, Linda A. Walsh wrote: >> Luca Berra wrote: >>>> I'm using a RAID 'chunk' size of 64k as suggested by the RAID >>>> documentation >>>> and am using 6 disks to create a RAID6, giving 4 units of >>>> data/stripe. Does >>> I suppose by raid you mean md, so i wonder what documentation you wer= e >>> looking at? >> --- >> Well, doc in 2 different raid controllers LSI and rocket raid both= >> suggest 64K as a unit size (forget, their exact term). Hardware raid and software raid are two entirely different things when it comes to optimization. >>> I think 64k might be small as a chunk size, depending on your array s= ize >>> you probably want a bigger size. >> --- >> Really? What are the trade offs? Array size well 6 disks and 4 >> of data. > ok, i trew the stone .. > First we have to consider usage scenarios, i.e. average read and averag= e > write size, large reads benefit from larger chunks, small writes with > too large chunks would still result on whole stripe Read-Modify-Write. >=20 > there were people on linux-raid ml doing benchmarks, and iirc using > chunks between 256k and 1m gave better average results That was me. The best results are with 256 or 512k chunk sizes. Above 512k you don't get any more benefit. --=20 Doug Ledford GPG KeyID: CFBFF194 http://people.redhat.com/dledford Infiniband specific RPMs available at http://people.redhat.com/dledford/Infiniband --------------enig5A1AB7CA1D2F6322B4DE4D92 Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) iEYEARECAAYFAkv+oJoACgkQg6WylM+/8ZSJPwCgqWYRd2amgNfahXoMq1muYIlB F1gAn25SSQ7kkj5GX5TwHpW5t+/KnaIA =HQ5W -----END PGP SIGNATURE----- --------------enig5A1AB7CA1D2F6322B4DE4D92--