From owner-freebsd-net@FreeBSD.ORG Fri Jan 20 15:04:21 2006 Return-Path: X-Original-To: freebsd-net@FreeBSD.org Delivered-To: freebsd-net@FreeBSD.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0915B16A41F for ; Fri, 20 Jan 2006 15:04:21 +0000 (GMT) (envelope-from glebius@FreeBSD.org) Received: from cell.sick.ru (cell.sick.ru [217.72.144.68]) by mx1.FreeBSD.org (Postfix) with ESMTP id C067E43D48 for ; Fri, 20 Jan 2006 15:04:19 +0000 (GMT) (envelope-from glebius@FreeBSD.org) Received: from cell.sick.ru (glebius@localhost [127.0.0.1]) by cell.sick.ru (8.13.3/8.13.3) with ESMTP id k0KF4GNT028052 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 20 Jan 2006 18:04:16 +0300 (MSK) (envelope-from glebius@FreeBSD.org) Received: (from glebius@localhost) by cell.sick.ru (8.13.3/8.13.1/Submit) id k0KF4GSI028051; Fri, 20 Jan 2006 18:04:16 +0300 (MSK) (envelope-from glebius@FreeBSD.org) X-Authentication-Warning: cell.sick.ru: glebius set sender to glebius@FreeBSD.org using -f Date: Fri, 20 Jan 2006 18:04:15 +0300 From: Gleb Smirnoff To: jhall@vandaliamo.net Message-ID: <20060120150415.GU83922@FreeBSD.org> Mail-Followup-To: Gleb Smirnoff , jhall@vandaliamo.net, freebsd-net@freebsd.org References: <1302.199.223.158.225.1137685960.squirrel@admintool.trueband.net> Mime-Version: 1.0 Content-Type: text/plain; charset=koi8-r Content-Disposition: inline In-Reply-To: <1302.199.223.158.225.1137685960.squirrel@admintool.trueband.net> User-Agent: Mutt/1.5.6i Cc: freebsd-net@FreeBSD.org Subject: Re: mpd 3.18 and FreeBSD 6.0 X-BeenThere: freebsd-net@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Networking and TCP/IP with FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 20 Jan 2006 15:04:21 -0000 On Thu, Jan 19, 2006 at 03:52:40PM -0000, jhall@vandaliamo.net wrote: j> Recently, I tried to upgrade my firewall from FreeBSD 4.8 to 6.0 and from j> mpd 3.17 to mpd 3.18. j> j> Initially, everything works. However, after a few hours (4 to 5) users j> start complaining about slow connection speeds, and eventually they are j> not able to connect to the Exchange server, or network drives shared over j> the WAN at all. If I put the FreeBSD 4.8 server and mpd 3.17 back on j> line, everything will start working properly almost immediately. To me, j> it sounds like a memory leak, or the operating system is running out of j> stack space. j> j> My network is configured in a star configuration. 5 offices connect to j> the corporate network using an mpd to mpd configuration. The corporate j> office has a dedicated T-1 and the other offices all have DSL connections. j> j> Nothing out of the ordinary is showing up in the mpd logs. And, nothing j> out of the ordinary is showing up in the syslog. j> j> In my mpd.conf file, I have 34 pptp connections defined, and at most 10 j> are in use at one time. The server is a Celeron processor with 128 MB of j> RAM. And, during peak usage times, I am showing approximately 1% usage on j> the processor. j> j> If needed, I would be happy to post my configuration files and log files. j> I have checked the log files around the times users report problems, and j> do not see anything out of the ordinary (no dropped connections, new j> connections, missed replies, etc). And, checking the T-1 usage at the j> time, approximately 500K of a full T-1 was in use. j> j> Any suggestions would be greatly appreciated. I guess you are using PPTP, although this word is absent in your mail. mpd can support different protocols, so reporting what you are actually using is important. The PPTP protocol has a terrible idea - windowing, that degradates tunnel performance a lot. In FreeBSD 4.8-RELEASE ng_pptpgre(4) didn't support windowing at all. In modern FreeBSD ng_pptpgre(4) supports windowing. In mpd windowing is turned on by default. This is done because some modern MS Windows versions will refuse to connect if windowing is disabled. I think, that if you disable windowing in mpd config, your performance will come back. -- Totus tuus, Glebius. GLEBIUS-RIPN GLEB-RIPE