From owner-freebsd-net Fri Oct 25 13:13:13 2002 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 8F52F37B401 for ; Fri, 25 Oct 2002 13:13:11 -0700 (PDT) Received: from power.doogles.com (power.doogles.com [209.15.149.130]) by mx1.FreeBSD.org (Postfix) with ESMTP id BB23E43E88 for ; Fri, 25 Oct 2002 13:13:09 -0700 (PDT) (envelope-from jyoung@power.doogles.com) Received: from localhost (jyoung@localhost) by power.doogles.com (8.11.6/8.11.6) with ESMTP id g9PKD6r02153; Fri, 25 Oct 2002 15:13:06 -0500 Date: Fri, 25 Oct 2002 15:13:06 -0500 (CDT) From: "Jason A. Young" To: Romain Kang Cc: Petri Helenius , Subject: Re: post-ifconfig delay causes ntpdate failure? In-Reply-To: <20021025192653.GA45730@kzsu.stanford.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, 25 Oct 2002, Romain Kang wrote: > On Fri, Oct 25, 2002 at 10:03:21PM +0300, Petri Helenius wrote: > > Are you sure that this is not caused by spanning tree delay on the ethernet > > switch you are probably connected to? > > Time for me to read up on STP bridging. However, I'd be a little > surprised that the delay would occur when both hosts are on the > same switch, and they were communicating immediately before the > client was rebooted (as in my tests). > > Whatever the cause, is there some method better than the ping loop > to determine if IP is actually getting out? > > Thanks, > Romain > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-net" in the body of the message > This is a pretty common problem; often it stops you from getting a DHCP lease because the DHCP client will time out while the switch still has you in the listening or learning states. It doesn't matter that the two devices are on the same switch. Whenever a switch port with spanning-tree enabled comes up, the switchport you're on will not accept any traffic from you for 30 seconds or so. It's likely that your ifconfig statement causes link to drop briefly. The solution for this is to not run spanning-tree on that switchport. Or, if you have a Cisco switch or another type of switch that supports skipping the listening and learning stages of spanning-tree for ports you know only workstations are connected to, you can enable that. On Cisco IOS-based switches this is 'spanning-tree portfast'; on Cisco Catalyst OS-based switches, use 'set port host X/Y' - this disables a few things, including spanning-tree. If you don't have access to the switch, you're basically going to have to put up with it. If it's causing a problem in your boot sequence, insert a 30-second sleep after ifconfig. -- Jason Young, CCIE #8607, MCSE Sr. Network Technician, WAN Technologies (314)817-0131 http://www.wantec.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message