From mboxrd@z Thu Jan 1 00:00:00 1970 From: Linus Torvalds Subject: Re: [GIT PULL] Ext3 latency fixes Date: Sat, 4 Apr 2009 15:19:19 -0700 (PDT) Message-ID: References: <1238742067-30814-1-git-send-email-tytso@mit.edu> <20090404135719.GA9812@mit.edu> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Cc: Jens Axboe , Linux Kernel Developers List , Ext4 Developers List To: Theodore Tso Return-path: Received: from smtp1.linux-foundation.org ([140.211.169.13]:45476 "EHLO smtp1.linux-foundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752607AbZDDWWE (ORCPT ); Sat, 4 Apr 2009 18:22:04 -0400 In-Reply-To: Sender: linux-ext4-owner@vger.kernel.org List-ID: On Sat, 4 Apr 2009, Linus Torvalds wrote: > > Hmm. I can certainly see a very noticeable improvement. .. and btw, just to put that into perspective - I ended up running with that repeated 'dd+sync+rm' in the background for quite a while, and the machine was in no way unusable. I don't think I saw a single "ooh, that feels really dead" case with anticipatory, although I saw several "bad stutter" cases. Of course, with the whole "overwrite" case, things are really much smoother, and I can hardly notice anything happening at all. Although I _do_ get an occasional 200-600ms fsync hickup, and I can tell those very clearly. But if that was the worst I'd ever see, I'd be a very happy person. It's a good goal to have, I suspect. Linus