From owner-freebsd-questions@FreeBSD.ORG Mon Oct 1 14:54:57 2007 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id A216016A41A for ; Mon, 1 Oct 2007 14:54:57 +0000 (UTC) (envelope-from bulkmail@diomedia.be) Received: from hoefnix.telenet-ops.be (hoefnix.telenet-ops.be [195.130.132.54]) by mx1.freebsd.org (Postfix) with ESMTP id 65CE613C46E for ; Mon, 1 Oct 2007 14:54:57 +0000 (UTC) (envelope-from bulkmail@diomedia.be) Received: from adicia.telenet-ops.be (adicia.telenet-ops.be [195.130.132.56]) by hoefnix.telenet-ops.be (Postfix) with ESMTP id 6587D9D1AD for ; Mon, 1 Oct 2007 16:35:55 +0200 (CEST) Received: from localhost (localhost.localdomain [127.0.0.1]) by adicia.telenet-ops.be (Postfix) with SMTP id AFBE12300A9 for ; Mon, 1 Oct 2007 16:35:53 +0200 (CEST) Received: from bsd.thuis.be (d5152F411.access.telenet.be [81.82.244.17]) by adicia.telenet-ops.be (Postfix) with ESMTP id 9CBFF23011E for ; Mon, 1 Oct 2007 16:35:53 +0200 (CEST) Message-ID: <470105B8.8080709@diomedia.be> Date: Mon, 01 Oct 2007 16:35:36 +0200 From: Bram Van Steenlandt User-Agent: Thunderbird 2.0.0.6 (X11/20070804) MIME-Version: 1.0 To: freebsd-questions@freebsd.org Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: roaming questions X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 01 Oct 2007 14:54:57 -0000 Hi list, I have setup a mobile pc to roam across our building. By reducing the dwell time and changing the channel list to only the channels I use roaming now works within ten seconds en sometimes within one or two seconds. The previous configuration was with fedora and there I was unable to get roaming time under 25 seconds. I do have one very annoying problem however and I have no idea how to solve it. The software uses psycopg (a python postgresql module) wich uses the standard system parameters for connections (At least that is what I think). -On fedora if the connection gets lost and it takes 30 seconds to remake a new connection operation is not interupted, after the 30 seconds you get the data you've been waiting for. -On freebsd however we get it a lot that the connection is "lost", you can easily start a new connection wich works fine, but the old connection you wore using stops working and the app. hangs (If I had to guess I would say that roaming works about 95% of the time and the connection is lost about 5% of the time). I also get a lot of IFDOWN IFUP messages but this seems normal to me. Is there some way to tune sysctl values so that it doesn't lose the connection or how should I proceed in debugging this problem ? Sorry about my poor english Tkanks in advance