Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756855AbZJJJwT (ORCPT ); Sat, 10 Oct 2009 05:52:19 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1756737AbZJJJwS (ORCPT ); Sat, 10 Oct 2009 05:52:18 -0400 Received: from mga06.intel.com ([134.134.136.21]:32928 "EHLO orsmga101.jf.intel.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1756345AbZJJJwR (ORCPT ); Sat, 10 Oct 2009 05:52:17 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.44,536,1249282800"; d="gz'50?scan'50,208,50";a="558316013" Subject: Re: fio rand read/write regression with 2.6.32-rc3 From: "Zhang, Yanmin" To: Jens Axboe Cc: LKML , Linus Torvalds In-Reply-To: <20091010080117.GN9228@kernel.dk> References: <1255153724.25078.68.camel@ymzhang> <20091010080117.GN9228@kernel.dk> Content-Type: multipart/mixed; boundary="=-EY2VlgPcUzHxOfsdjloT" Date: Sat, 10 Oct 2009 17:52:13 +0800 Message-Id: <1255168333.25078.82.camel@ymzhang> Mime-Version: 1.0 X-Mailer: Evolution 2.22.1 (2.22.1-2.fc9) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3610 Lines: 84 --=-EY2VlgPcUzHxOfsdjloT Content-Type: text/plain Content-Transfer-Encoding: 7bit On Sat, 2009-10-10 at 10:01 +0200, Jens Axboe wrote: > On Sat, Oct 10 2009, Zhang, Yanmin wrote: > > Comparing with 2.6.23-rc1's result, fio rand read write has regression > > on my 2*4 core stoakley machine (8GB memory) with a JBOD of 12 disks. > > > > Every disk has 8 1-GB files. Start 8 sub-processes per disk and every > > process random chooses a file on the disk to do 36 times of file read or > > write on the file and then choose another file. > > > > > > fio_mmap_rand_read_4k regresion is about 35%. > > Heh, I seem to recollect I told Linus that this would cost is 30-40% > performance. So not totally crazy. > > So yes, this isn't hugely unexpected. If you send me your fio job files, > I'll try and see what I can do about it. See the attachment. > > > Bisect down to patch: > > 1d2235152dc745c6d94bedb550fea84cffdbf768 is first bad commit > > commit 1d2235152dc745c6d94bedb550fea84cffdbf768 > > Author: Jens Axboe > > Date: Fri Oct 2 19:27:04 2009 +0200 > > > > cfq-iosched: add a knob for desktop interactiveness > > > > > > After I revert the patch against 2.6.23-rc3, fio_mmap_rand_read_4k regression > > disappears. > > > > fio_mmap_randrw_4k has less than 20% regression and reverting the patch > > could restore performance. > > > > On another stoakley machine with another JBOD, we see the similiar regression. > > > > ffsb rand read/write has a bigger regression and reverting the patch > > could restore performance. > > > > Yanmin > > > > > --=-EY2VlgPcUzHxOfsdjloT Content-Disposition: attachment; filename=job_file_st01_fio_mmap_randread_4k.gz Content-Type: application/x-gzip; name=job_file_st01_fio_mmap_randread_4k.gz Content-Transfer-Encoding: base64 H4sICJ5Y0EoAA2pvYl9maWxlX3N0MDFfZmlvX21tYXBfcmFuZHJlYWRfNGsApdrLbtswFIThvV4m pu4u4HUfoigMJZZdJbZkyEqL9OnrXmbozurwaEcggT/iB0Yrfjmdp+fu/DU7DHP/suw22TD142kY +93l0l2z2/Cz37Wfs+fbrnzLxvfL63Q/huw8Tdfbrsrm93EZLv2u2Gyy0zy9X/dzf53mZRhP2TB+ 787DoVv6+8/+/f1p/tg9Xcbl6bZcn47DdP9jlx2Hc7+/9fP34aXfLx/Xfjd342G6fCrqLPtyB/e3 w3PY396fN1+z29LNy6E/dx/3H51//PnXue8Of35l7O5X+f2bm6dj+PTvkONQ4FDiUOFQ49Dg0Aod EminiEMQOk+gnSJbCV0k0E6RrYQuE2inyFZCVwm0U2QroesE2imyldBNAu0U2Yr0S9quA8QAMUAM EAPEADGgVXjc9Uvarv0iDkFo8679IlsJbd61X2Qroc279otsJbR5136RrYQ279ovspXQ5l37RbYi fUjbdQ4xh5hDzCHmEHOIOVrlj7s+pO3aL+IQhDbv2i+yldDmXftFthLavGu/yFZCm3ftF9lKaPOu /SJbCW3etV9kK9J92q4LiAXEAmIBsYBYQCzQqnjcdZ+2a7+IQxDavGu/yFZCm3ftF9lKaPOu/SJb CW3etV9kK6HNu/aLbCW0edd+ka1IH9N2XUIsIZYQS4glxBJiiVbl466Pabv2izgEoc279otsJbR5 136RrYQ279ovspXQ5l37RbYS2rxrv8hWQpt37RfZivQpbdcVxApiBbGCWEGsIFZoVT3u+pS2a7+I QxDavGu/yFZCm3ftF9lKaPOu/SJbCW3etV9kK6HNu/aLbCW0edd+ka1If0vbdQ2xhlhDrCHWEGuI NVrVj7v+lrZrv4hDENq8a7/IVkKbd+0X2Upo8679IlsJbd61X2Qroc279otsJbR5136RrUgPabtu IDYQG4gNxAZiA7FBq+Zx10Parv0iDkFo8679IlsJbd61X2Qroc279otsJbR5136RrYQ279ovspXQ 5l37RbYi/Zq26xZiC7GF2EJsIbYQW7RqH3f9mrZrv4hDENq8a7/IVkKbd+0X2Upo8679IlsJbd61 X2Qroc279otsJbR5136RrUi/pe16C3ELcQtxC3ELcQtxi1bbx12/pe3aL+IQhDbv2i+yldDmXftF thLavGu/yFZCm3ftF9lKaPOu/SJbCW3etV9kK9LnxHcpfAoT+BYm8DFM4GuYwOcwge9hAp/yhP/e nZ0T36eshXkKcgX7O5W1cCwoV7C/V1kLx4JyBfu7lbVwLChXsL9fWQvHgnIF+zuWtXAsKFewv2dZ C8eCvMIl8bsQn9LEtzTxMU18TROf08T3NPEp0H/v1i6J34W1ME9BrmD/LqyFY0G5gv27sBaOBeUK 9u/CWjgWlCvYvwtr4VhQrmD/LqyFY0G5gv27sBaOBbPsF2PElBJBLwAA --=-EY2VlgPcUzHxOfsdjloT-- -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/