From: Florian Weimer Subject: Re: Bug#605009: serious performance regression with ext4 Date: Mon, 29 Nov 2010 16:27:12 +0000 Message-ID: <82oc98f63j.fsf@mid.bfk.de> References: <20101126093257.23480.86900.reportbug@pluto.milchstrasse.xx> <20101129072930.GA7213@burratino> <20101129144436.GT2767@thunk.org> <201011291618.25084.bernd.schubert@fastmail.fm> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: "Ted Ts'o" , Jonathan Nieder , linux-ext4@vger.kernel.org To: Bernd Schubert Return-path: Received: from mx01.bfk.de ([193.227.124.2]:35852 "EHLO mx01.bfk.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755072Ab0K2QeU convert rfc822-to-8bit (ORCPT ); Mon, 29 Nov 2010 11:34:20 -0500 In-Reply-To: <201011291618.25084.bernd.schubert@fastmail.fm> (Bernd Schubert's message of "Mon\, 29 Nov 2010 16\:18\:24 +0100") Sender: linux-ext4-owner@vger.kernel.org List-ID: * Bernd Schubert: > Wouldn't it make sense to modify ext4 or even the vfs to do that on c= lose()=20 > itself? Most applications expect the file to be on disk after a close= anyway=20 > and I also don't see a good reason why one should delay a disk write-= back=20 > after close any longer (well, there are exeption if the application i= s broken,=20 > for example such as ha-logd used by pacemaker, which did for each lin= e of logs=20 > an open, seek, write, flush, close sequence..., but at least we have = fixed=20 > that in -hg now). If you use Oracle Berkeley DB in a process-based fashion, it is crucial for decent performance that the memory-mapped file containing the cache is not flushed to disk when the database environment is closed prior to process termination. Perhaps flushing could be delayed until the last open file handle is gone. In any case, it's a pretty drastic change, which should probably be tunable with a (generic) mount option. --=20 =46lorian Weimer BFK edv-consulting GmbH http://www.bfk.de/ Kriegsstra=DFe 100 tel: +49-721-96201-1 D-76133 Karlsruhe fax: +49-721-96201-99 -- 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