From owner-freebsd-net@FreeBSD.ORG Sun Jun 20 22:08:15 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 3C27B16A4CE for ; Sun, 20 Jun 2004 22:08:15 +0000 (GMT) Received: from hanoi.cronyx.ru (hanoi.cronyx.ru [144.206.181.53]) by mx1.FreeBSD.org (Postfix) with ESMTP id 627F843D4C for ; Sun, 20 Jun 2004 22:08:14 +0000 (GMT) (envelope-from rik@cronyx.ru) Received: (from root@localhost) by hanoi.cronyx.ru id i5KM22CW056596 for net@freebsd.org.checked; (8.12.8/vak/2.1) Mon, 21 Jun 2004 02:02:02 +0400 (MSD) (envelope-from rik@cronyx.ru) Received: from cronyx.ru (rik.cronyx.ru [172.22.4.1]) by hanoi.cronyx.ru with ESMTP id i5KM0USr056542; (8.12.8/vak/2.1) Mon, 21 Jun 2004 02:00:31 +0400 (MSD) (envelope-from rik@cronyx.ru) Message-ID: <40D60791.8080200@cronyx.ru> Date: Mon, 21 Jun 2004 01:54:25 +0400 From: Roman Kurakin User-Agent: Mozilla/5.0 (X11; U; Linux i686; ru-RU; rv:1.2.1) Gecko/20030426 X-Accept-Language: ru-ru, en MIME-Version: 1.0 To: Joerg Wunsch References: <40D4C79B.2050400@cronyx.ru> <20040620232936.D13428@uriah.heep.sax.de> In-Reply-To: <20040620232936.D13428@uriah.heep.sax.de> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit cc: net@freebsd.org Subject: Re: if_sppp 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: Sun, 20 Jun 2004 22:08:15 -0000 Joerg Wunsch: >As Roman Kurakin wrote: > > > >>Problem: >> If we have max_failure < MAXALIVECNT*5 we will >>send conf-rej for magic. >> >> >>Solution: >> Loopback could be treated as a special case and >>thus we may not count it as a failure. >> >> > >Can you explain a little more, please? The patch is simple enough, >yes, but offhand I don't know what's the actual problem resulting out >of the above situation. > > Ok. If we have loopback, we will continue to try with some new value of magic. In every try we will increase both fail_count and loopback count. After every try we will check both counters if they exceed appropriate max value and will take some action if one of them does. But since max_failue is 10 and max for loopback counter is 15 we will have only special action after axceed of fail_counter. This in turn means that we will reject magic and since we are loopbacked we will recieve it and turn off magic. After next attempt we will take any magic (since we turn it off) and will move from lcp to ipcp or whatever else. I used to use this for driver testings, since state machine becomes crazy and I get very good traffic. ;-) But this definitely should be fixed. I think the best way to fix this is to treate loopack in different way from other failures. And since loopbacks have its own counter they should be counted as failures. rik