From: Yongqiang Yang Subject: Re: [PATCH 0/2] ext3: Reduce calling ext3_mark_inode_dirty() for speedup Date: Fri, 3 Feb 2012 21:28:02 +0800 Message-ID: References: <4F2657AD.9060901@sx.jp.nec.com> <97F0BDAD-B6D8-4246-B790-E269025F4A7D@dilger.ca> <4F28F962.2040209@sx.jp.nec.com> <4F2B919D.1030307@sx.jp.nec.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Andreas Dilger , ext4 , Jan Kara To: Kazuya Mio Return-path: Received: from mail-tul01m020-f174.google.com ([209.85.214.174]:56209 "EHLO mail-tul01m020-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755024Ab2BCN2E convert rfc822-to-8bit (ORCPT ); Fri, 3 Feb 2012 08:28:04 -0500 Received: by obcva7 with SMTP id va7so4236802obc.19 for ; Fri, 03 Feb 2012 05:28:02 -0800 (PST) In-Reply-To: <4F2B919D.1030307@sx.jp.nec.com> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Fri, Feb 3, 2012 at 3:49 PM, Kazuya Mio wrote: > 2012/02/03 7:36, Andreas Dilger wrote: >>> >>> =A0filesystem =A0 =A0 =A0 =A0time(sec) =A0call extX_mark_inode_dirt= y(times) >>> =A0--- >>> =A0ext3 =A0 =A0 =A0 =A0 =A0 =A0 =A0220.5 =A0 =A0 =A050,338,104 >>> =A0ext3 (patched) =A0 =A0196.3 =A0 =A0 =A025,169,658 >>> =A0ext4 (*1) =A0 =A0 =A0 =A0 190.3 =A0 =A0 =A028,465,799 >>> =A0ext4 (*2) =A0 =A0 =A0 =A0 201.5 =A0 =A0 =A027,963,473 >>> =A0ext4 (default) =A0 =A0223.3 =A0 =A0 =A014,026,118 >>> >>> =A0*1 disable ext4-specific options (delalloc, extent, and so on) >>> =A0*2 disable only delalloc option >> >> This shows that ext4 with extents+delalloc is _slower_ than ext3, wh= ich >> is very strange. =A0In other similar tests of write performance (see > > > One more thing is that ext4+delalloc is slower than ext4+nodelalloc. And according to the data, maybe ext4+extent is also slower than ext4+n= oextent. What's the size of the fs? and what kind of the tested device? Yongqiang. > > >> http://downloads.linux.hp.com/~enw/ext4/3.2/large_file_creates.html, >> showing multi-threaded 1GB file writes) ext4 is much faster than ext= 3. > > > I guess write buffer size of my test is different from ffsb's one. > My test calls write systemcall every time one block is allocated, > so it is close to the stress test I think. > > >> Looking at your original email, is ext4 being tested on a RHEL 5.5 >> (2.6.18) kernel, or a more recent kernel? =A0It would be more useful >> to run this on a more modern kernel, since the ext4 code backported >> to RHEL5 was barely supporting delalloc at all, if I remember correc= tly. > > > I tested on the recent kernel (3.3-rc1). > I also tested on RHEL5.5, and its result showed that ext3 was much sl= ower > than > the recent kernel's one. > > =A0filesystem =A0 =A0 =A0 =A0time(sec) > =A0--- > =A0ext3(RHEL5.5) =A0 =A0 438.6 > =A0ext3(3.3-rc1) =A0 =A0 220.5 > > Regards, > Kazuya Mio > > -- > To unsubscribe from this list: send the line "unsubscribe linux-ext4"= in > the body of a message to majordomo@vger.kernel.org > More majordomo info at =A0http://vger.kernel.org/majordomo-info.html --=20 Best Wishes Yongqiang Yang -- To unsubscribe from this list: send the line "unsubscribe linux-ext4" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html