From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Beverley Subject: Re: QoS weirdness : HTB accuracy Date: Thu, 10 Jun 2010 22:22:28 +0100 Message-ID: <1276204948.1403.13.camel@andybev> References: <067c83163988908ef546d7ff7f560a17@localhost> <4BF32BB0.7010700@redfish-solutions.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: "Philip A. Prindeville" , Netdev , netfilter To: Julien Vehent Return-path: In-Reply-To: Sender: netfilter-owner@vger.kernel.org List-Id: netdev.vger.kernel.org > > Sorry for the late response: could this be an "aliasing" issue caused > > by sampling intervals (granularity)? > > > > I was, in fact, an error in my ruleset. I had put the 'linklayer atm' at > both the branch and leaf levels, so the overhead was computed twice, > creating those holes in the bandwidth. I am seeing similar behaviour with my setup. Am I making the same mistake? A subset of my rules is as follows: tc qdisc add dev ppp0 root handle 1: htb r2q 1 tc class add dev ppp0 parent 1: classid 1:1 htb \ rate ${DOWNLINK}kbit ceil ${DOWNLINK}kbit \ overhead $overhead linklayer atm <------- Here tc class add dev ppp0 parent 1:1 classid 1:10 htb \ rate 612kbit ceil 612kbit prio 0 \ overhead $overhead linklayer atm <------- And here tc qdisc add dev ppp0 parent 1:10 handle 4210: \ sfq perturb 10 limit 50 tc filter add dev ppp0 parent 1:0 protocol ip \ prio 10 handle 10 fw flowid 1:10 Thanks, Andy