From: Roman Mamedov Subject: Re: Filesystem writes on RAID5 too slow Date: Tue, 19 Nov 2013 00:41:40 +0600 Message-ID: <20131119004140.3fea96df@natsu> References: Mime-Version: 1.0 Content-Type: multipart/signed; micalg=PGP-SHA1; boundary="Sig_/x4nyDRcD=BOV5KkYrQUcxCe"; protocol="application/pgp-signature" Cc: "Kernel.org-Linux-RAID" , "Kernel.org-Linux-XFS" , "Kernel.org-Linux-EXT4" To: Martin Boutin Return-path: In-Reply-To: Sender: linux-raid-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org --Sig_/x4nyDRcD=BOV5KkYrQUcxCe Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable On Mon, 18 Nov 2013 11:02:15 -0500 Martin Boutin wrote: > I have the md0 raid device formatted as ext3 with a 4k block size, and > stride and stripes correctly chosen to match the raid chunk size, that > is, stride=3D128,stripe-width=3D256. What is your stripe cache size? http://peterkieser.com/2009/11/29/raid-mdraid-stripe_cache_size-vs-write-tr= ansfer/ > The command line for measuring filesystem read and write speeds was: >=20 > $ dd if=3D/tmp/diskmnt/filerd.zero of=3D/dev/null bs=3D1M count=3D1000 if= lag=3Ddirect > $ dd if=3D/dev/zero of=3D/tmp/diskmnt/filewr.zero bs=3D1M count=3D1000 of= lag=3Ddirect Try testing with "fdatasync" instead of "direct" here. --=20 With respect, Roman --Sig_/x4nyDRcD=BOV5KkYrQUcxCe Content-Type: application/pgp-signature; name=signature.asc Content-Disposition: attachment; filename=signature.asc -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) iEYEARECAAYFAlKKX2QACgkQTLKSvz+PZwh1mwCgjuGOIMIxOqMQRP79r1fKVx+f 1f0AnR9v0sYECiXgCFZoXoUzhVdBmPwO =kJZ2 -----END PGP SIGNATURE----- --Sig_/x4nyDRcD=BOV5KkYrQUcxCe--