From owner-freebsd-net@FreeBSD.ORG Thu Feb 22 16:44:29 2007 Return-Path: X-Original-To: freebsd-net@freebsd.org Delivered-To: freebsd-net@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id C528416A403 for ; Thu, 22 Feb 2007 16:44:29 +0000 (UTC) (envelope-from tarkhil@webmail.sub.ru) Received: from mail.sub.ru (mail.sub.ru [88.212.205.2]) by mx1.freebsd.org (Postfix) with SMTP id 0ACCD13C442 for ; Thu, 22 Feb 2007 16:44:28 +0000 (UTC) (envelope-from tarkhil@webmail.sub.ru) Received: (qmail 52883 invoked by uid 0); 22 Feb 2007 19:50:33 +0300 Received: from unknown (HELO localhost) (88.212.205.2) by techno.sub.ru with SMTP; 22 Feb 2007 16:50:33 -0000 X-Virus-Scanned: by amavisd-new at mail.sub.ru Received: from unknown ([88.212.205.2]) by localhost (mail-new.sub.ru [88.212.205.2]) (amavisd-new, port 10024) with SMTP id 6lV9XLCRpmI4; Thu, 22 Feb 2007 19:50:25 +0300 (MSK) Received: from unknown (HELO ?192.168.139.47?) (tarkhil%sub.ru@213.247.139.99) by techno.sub.ru with SMTP; 22 Feb 2007 16:50:25 -0000 Message-ID: <45DDC852.9060503@webmail.sub.ru> Date: Thu, 22 Feb 2007 19:44:02 +0300 From: Alex Povolotsky User-Agent: Thunderbird 1.5.0.9 (X11/20070104) MIME-Version: 1.0 To: Alexander Motin References: <1172056982.00691952.1172043601@10.7.7.3> <45DC2FC1.3030408@alkar.net> <45DC3718.7010804@webmail.sub.ru> <45DC39BC.5020707@alkar.net> <45DC44AF.30005@webmail.sub.ru> <45DC4640.8030309@alkar.net> <45DDBBB3.3020900@webmail.sub.ru> <45DDC091.3060903@alkar.net> In-Reply-To: <45DDC091.3060903@alkar.net> Content-Type: text/plain; charset=KOI8-R; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-net@freebsd.org Subject: Re: mpd success stories, anyone? 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: Thu, 22 Feb 2007 16:44:29 -0000 Alexander Motin wrote: > Alex Povolotsky wrote: >> Thanks, it seems to be more or less stable; however, throughput is >> quite little, lots of packets lost and "No buffer space available" on >> attempt to ping VPN addresses (only VPN is affected). > > Have you tried to disable PPTP windowing in mpd config? ENOBUFS is the > errno used by ng_pptp node's windowing code when outgoing window is > full. It is not related to any system tunables. Maximum window size in > current ng_pptp is 16 packets. It can be small for LFNs and can reduce > speed. > >> I guess I should tune some kernel tunable, but what specific one? > > I can recomend you to set > net.graph.maxdgram=524288 > net.graph.recvspace=524288 > to make 'ngctl list' command work, but it is not critical. > Whoops! After disabling windowing and setting net.graph's, mpd4 refuses to work Feb 22 19:41:43 gw mpd: can't create socket node: No buffer space available Feb 22 19:41:43 gw mpd: pptp0: killing connection with 172.23.115.234 1735 Feb 22 19:41:44 gw mpd: PPTP connection from 172.23.114.214 2955 Feb 22 19:41:44 gw mpd: pptp0: attached to connection with 172.23.114.214 2955 Feb 22 19:41:44 gw mpd: [pptp7] Accepting PPTP connection Feb 22 19:41:44 gw mpd: [pptp7] opening link "pptp7"... Feb 22 19:41:44 gw mpd: [pptp7] link: OPEN event Feb 22 19:41:44 gw mpd: [pptp7] LCP: Open event Feb 22 19:41:44 gw mpd: [pptp7] LCP: state change Initial --> Starting Feb 22 19:41:44 gw mpd: [pptp7] LCP: LayerStart Feb 22 19:41:44 gw mpd: [pptp7] attaching to peer's outgoing call Feb 22 19:41:44 gw mpd: [pptp7] can't attach pptpgre node: Bad file descriptor Feb 22 19:41:44 gw mpd: pptp0-0: killing channel Feb 22 19:41:44 gw mpd: [pptp7] PPTP call cancelled in state CONNECTING Feb 22 19:41:44 gw mpd: pptp0: closing connection with 172.23.114.214 2955 Feb 22 19:41:44 gw mpd: [pptp7] can't shutdown "bypass.link0": Bad file descript or Feb 22 19:41:44 gw mpd: [pptp7] link: DOWN event Feb 22 19:41:44 gw mpd: [pptp7] LCP: Close event Feb 22 19:41:44 gw mpd: [pptp7] LCP: state change Starting --> Initial Feb 22 19:41:44 gw mpd: [pptp7] LCP: LayerFinish and no ng interfaces ever created lowering both tunables to 128000 solved the problem, will look more. Alex.