From owner-freebsd-net@FreeBSD.ORG Mon Nov 22 13:18:55 2004 Return-Path: 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 B9A4816A4CE for ; Mon, 22 Nov 2004 13:18:55 +0000 (GMT) Received: from insomnia.benzedrine.cx (insomnia.benzedrine.cx [62.65.145.30]) by mx1.FreeBSD.org (Postfix) with ESMTP id A567243D5F for ; Mon, 22 Nov 2004 13:18:54 +0000 (GMT) (envelope-from dhartmei@insomnia.benzedrine.cx) Received: from insomnia.benzedrine.cx (dhartmei@localhost [127.0.0.1]) iAMDIr5x032491 (version=TLSv1/SSLv3 cipher=DHE-DSS-AES256-SHA bits=256 verify=NO); Mon, 22 Nov 2004 14:18:53 +0100 (MET) Received: (from dhartmei@localhost) by insomnia.benzedrine.cx (8.13.1/8.12.10/Submit) id iAMDIqD4001279; Mon, 22 Nov 2004 14:18:52 +0100 (MET) Date: Mon, 22 Nov 2004 14:18:52 +0100 From: Daniel Hartmeier To: Pawel Worach Message-ID: <20041122131851.GA14128@insomnia.benzedrine.cx> References: <419EBE2E.9080108@telia.com> <200411212028.15015.max@love2party.net> <41A0F1EF.7020804@telia.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <41A0F1EF.7020804@telia.com> User-Agent: Mutt/1.4.1i cc: Max Laier cc: freebsd-net@freebsd.org Subject: Re: SACK (and PF) wierdness X-BeenThere: freebsd-net@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Networking and TCP/IP with FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 22 Nov 2004 13:18:55 -0000 Pawel, could you provide a tcpdump -nvvvSXpi for an entire connection, from handshake to the point where it stalls? Please include the corresponding 'BAD state'/'State failure' messages and output of pfctl -vvss related to the connection (ideally all for the same connection, so timestamps are comparable). If the output is too long to post here, put it on a web server and post the URL, or mail the dump to Max and me privately. It does look like one peer is violating the other's window by sending too much data before acknowledgement. The state has picked up on the window scaling, so we need to see the entire TCP connection to determine whether the packet is dropped correctly or not. Daniel