From: NeilBrown Subject: Re: ext4 write performance regression in 3.6-rc1 on RAID0/5 Date: Wed, 22 Aug 2012 16:00:25 +1000 Message-ID: <20120822160025.272188d1@notabene.brown> References: <20120816024654.GB3781@thunk.org> <20120816111051.GA16036@localhost> <20120816152513.GA31346@thunk.org> <20120817060915.GB28786@localhost> <20120817134039.GB11439@thunk.org> <20120817142526.GA1059@localhost> <20120822035702.GF2570@yliu-dev.sh.intel.com> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=PGP-SHA1; boundary="Sig_/doLWNd9A289=bYVeFpiGDTq"; protocol="application/pgp-signature" Cc: Fengguang Wu , Li Shaohua , Theodore Ts'o , Marti Raudsepp , Kernel hackers , ext4 hackers , maze@google.com, "Shi, Alex" , linux-fsdevel@vger.kernel.org, linux RAID To: Yuanhan Liu Return-path: In-Reply-To: <20120822035702.GF2570@yliu-dev.sh.intel.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org --Sig_/doLWNd9A289=bYVeFpiGDTq Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable On Wed, 22 Aug 2012 11:57:02 +0800 Yuanhan Liu wrote: > =20 > -#define NR_STRIPES 256 > +#define NR_STRIPES 1024 Changing one magic number into another magic number might help your case, b= ut it not really a general solution. Possibly making sure that max_nr_stripes is at least some multiple of the chunk size might make sense, but I wouldn't want to see a very large multip= le. I thing the problems with RAID5 are deeper than that. Hopefully I'll figure out exactly what the best fix is soon - I'm trying to look into it. I don't think the size of the cache is a big part of the solution. I think correct scheduling of IO is the real answer. Thanks, NeilBrown --Sig_/doLWNd9A289=bYVeFpiGDTq Content-Type: application/pgp-signature; name=signature.asc Content-Disposition: attachment; filename=signature.asc -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.18 (GNU/Linux) iQIVAwUBUDR1gznsnt1WYoG5AQIvMBAAnwqYNlar29oqFwn2sT2XzxABU10uA6SW px7PJihjVfckygzmPTIibCXb7CdDnKUflXvdTi6YP+ZUEjAmbKVg4OQwkkpZcFoz Xevjdk+SyvSAYZFpiZFqgTpWV8XySW1HO8hfC8aagB3mEKYwd7pLaT0SF5Arer9P e91hUF3SR4Kvytoig45Ucf6KiXHBghbJ6oU2ARGFKWLv4yrOdR/WKs8z9nvGmDHL HMvXY2FONYtxevJ1nWPUp1/gk0cbkpfd/qrZhGopKok/ezQ6CDm4/WpQZZ04dYF9 cQ0ML57fJT8gR9deSIHKyiu89xVgItFoa5YQpa7gZMyk/9Cr1bierN4OCe8YDV2E 9gydeFhw1kbWDUEotPhr5Tk2qazjZg1Kt26Ea0zxBetJNX1jwTPaRnsPFyUMsGCj WUYQ3pBLLUrcec3kJcVz77+HN78AtvoxDE/ZKC35OrlAU0Lh5T3qJ8dThx31jM8C wAqABoqPDeqBUKCe+OA7SRxugvXrpolZ5JH+mGywl3Cp1PHQ7J4G0tHCImHLme+r 9gMrj6uCVacof/W+g+of0n2fKYC1AekxrRcyRXmzZi4AEla06kCy+DOv0ipb99if FaJORbwfZjcd3dOKp+jZg+6UFKWiZ+Clhs5muS2ejkVlagcI/ByP6W53R6EsA8Ds Bei+hDYJNEw= =6+qx -----END PGP SIGNATURE----- --Sig_/doLWNd9A289=bYVeFpiGDTq--