From owner-freebsd-current@FreeBSD.ORG Tue Nov 16 17:46:18 2004 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 9618E16A4CE for ; Tue, 16 Nov 2004 17:46:18 +0000 (GMT) Received: from nerve.riss-telecom.ru (nerve.riss-telecom.ru [80.66.65.3]) by mx1.FreeBSD.org (Postfix) with ESMTP id 8234B43D49 for ; Tue, 16 Nov 2004 17:46:17 +0000 (GMT) (envelope-from frol@nerve.riss-telecom.ru) Received: from nerve.riss-telecom.ru (nqnpulf0mxxrzz00@localhost [127.0.0.1]) by nerve.riss-telecom.ru (8.13.1/8.13.1) with ESMTP id iAGHjb7l017791 for ; Tue, 16 Nov 2004 23:45:37 +0600 (NOVT) (envelope-from frol@nerve.riss-telecom.ru) Received: (from frol@localhost) by nerve.riss-telecom.ru (8.13.1/8.13.1/Submit) id iAGHjaUm017790 for freebsd-current@freebsd.org; Tue, 16 Nov 2004 23:45:36 +0600 (NOVT) (envelope-from frol) Date: Tue, 16 Nov 2004 23:45:36 +0600 From: Dmitry Frolov To: freebsd-current@freebsd.org Message-ID: <20041116174536.GA17279@nerve.riss-telecom.ru> Mail-Followup-To: freebsd-current@freebsd.org References: <200411160916.47131.Emanuel.Strobl@gmx.net> <39261.1100593136@critter.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <39261.1100593136@critter.freebsd.dk> Organization: RISS-Telecom, JSC X-PGP-Fingerprint: 5232 98E7 596E 21C2 52B5 FCAE 8088 3F87 88BC 27B0 User-Agent: Mutt/1.5.6i Subject: Re: TEST: phk_bufwork patchfile X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 16 Nov 2004 17:46:18 -0000 * Poul-Henning Kamp [16.11.2004 14:19]: > >My intents are to make storage available for jails via regular files which > >is at the moment no option since performance gets horrible. If I understand > >things right then this framework should improove performance dramatically > >when e. g. copying /sys to /vnode_backed_md0. > > Not particularly. Performance issue with a file backed filesystems is a major showstopper for me too. Since I'm not a kernel hacker, I'm currenly waiting for someone to fix this. If I understand issue correctly, fixing it requires much work on syncer (probably some sort of "layering"?). Does anybody working or planning to work on this? wbr&w, dmitry. -- Dmitry Frolov RISS-Telecom Network, Novosibirsk, Russia 66415911@ICQ, +7 3832 NO WA1T, DVF-RIPE