From owner-freebsd-pf@FreeBSD.ORG Wed May 29 16:13:21 2013 Return-Path: Delivered-To: freebsd-pf@freebsd.org Received: from mx1.freebsd.org (mx1.FreeBSD.org [8.8.178.115]) by hub.freebsd.org (Postfix) with ESMTP id B29B9633 for ; Wed, 29 May 2013 16:13:21 +0000 (UTC) (envelope-from emz@norma.perm.ru) Received: from elf.hq.norma.perm.ru (unknown [IPv6:2001:470:1f09:14c0::2]) by mx1.freebsd.org (Postfix) with ESMTP id 3F2EE6DD for ; Wed, 29 May 2013 16:13:20 +0000 (UTC) Received: from [192.168.248.34] ([192.168.248.34]) by elf.hq.norma.perm.ru (8.14.5/8.14.5) with ESMTP id r4TGDEFE052049 (version=TLSv1/SSLv3 cipher=DHE-RSA-CAMELLIA256-SHA bits=256 verify=NO) for ; Wed, 29 May 2013 22:13:16 +0600 (YEKT) (envelope-from emz@norma.perm.ru) Message-ID: <51A62915.1060505@norma.perm.ru> Date: Wed, 29 May 2013 22:13:09 +0600 From: "Eugene M. Zheganin" User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:17.0) Gecko/20130509 Thunderbird/17.0.6 MIME-Version: 1.0 To: freebsd-pf@freebsd.org Subject: pf/altq and default queue Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.2.7 (elf.hq.norma.perm.ru [192.168.3.10]); Wed, 29 May 2013 22:13:16 +0600 (YEKT) X-Spam-Status: No hits=-101.0 bayes=0.5 testhits ALL_TRUSTED=-1, USER_IN_WHITELIST=-100 autolearn=unavailable version=3.3.2 X-Spam-Checker-Version: SpamAssassin 3.3.2 (2011-06-06) on elf.hq.norma.perm.ru X-BeenThere: freebsd-pf@freebsd.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: "Technical discussion and general questions about packet filter \(pf\)" List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 29 May 2013 16:13:21 -0000 Hi. I have an 8.1-RELEASE (yup, a bit outdated). Today I saw a situation when I observe a speed about 90 MBit/sec on an interface where altq is enabled, but root queue shows only half of that amount. Here comes the output: config: altq on em0 cbq bandwidth 100Mb queue { domru, internal } queue domru bandwidth 30Mb { domru_voip, domru_term, domru_standard } queue domru_voip bandwidth 2Mb priority 7 cbq(borrow) queue domru_term bandwidth 2Mb priority 7 cbq(borrow) queue domru_standard bandwidth 25Mb priority 4 cbq(red) queue internal bandwidth 70Mb cbq (default) pfctl -vvvs queue: queue root_em0 on em0 bandwidth 100Mb priority 0 cbq( wrr root ) {domru, internal} [ pkts: 5640756 bytes: 4902213386 dropped pkts: 0 bytes: 0 ] [ qlength: 0/ 50 borrows: 0 suspends: 0 ] [ measured: 8895.4 packets/s, 50.82Mb/s ] queue domru on em0 bandwidth 30Mb {domru_voip, domru_term, domru_standard} [ pkts: 0 bytes: 0 dropped pkts: 0 bytes: 0 ] [ qlength: 0/ 50 borrows: 0 suspends: 0 ] [ measured: 0.0 packets/s, 0 b/s ] queue domru_voip on em0 bandwidth 2Mb priority 7 cbq( borrow ) [ pkts: 23 bytes: 12029 dropped pkts: 0 bytes: 0 ] [ qlength: 0/ 50 borrows: 0 suspends: 0 ] [ measured: 0.0 packets/s, 170.36 b/s ] queue domru_term on em0 bandwidth 2Mb priority 7 cbq( borrow ) [ pkts: 18257 bytes: 12440795 dropped pkts: 27 bytes: 8306 ] [ qlength: 23/ 50 borrows: 9196 suspends: 643 ] [ measured: 132.6 packets/s, 977.86Kb/s ] queue domru_standard on em0 bandwidth 25Mb priority 4 cbq( red ) [ pkts: 1853521 bytes: 1561628960 dropped pkts: 46210 bytes: 36121115 ] [ qlength: 0/ 50 borrows: 0 suspends: 91436 ] [ measured: 2828.0 packets/s, 18.86Mb/s ] queue internal on em0 bandwidth 70Mb cbq( default ) [ pkts: 3768965 bytes: 3328140798 dropped pkts: 1887 bytes: 1185226 ] [ qlength: 0/ 50 borrows: 0 suspends: 84630 ] [ measured: 5934.9 packets/s, 30.98Mb/s ] netstat -I em0 1: 4616 0 0 4622558 3899 25 2795894 0 4470 0 0 4381202 4245 43 3324893 0 4589 0 0 4189727 4759 24 3797872 0 4231 0 0 4367365 3766 25 2687472 0 4838 0 0 4860147 4188 26 3022562 0 5025 0 0 5168430 4418 69 3226557 0 6379 0 0 7429234 5361 11 2663720 0 <=== wget started 8264 0 0 10792601 5921 17 1620189 0 8449 0 0 11165433 6006 25 1262430 0 8736 0 0 11428083 6010 37 1007035 0 8513 0 0 10751876 6107 33 1635660 0 8511 0 0 10925151 6083 40 1474535 0 8244 0 0 10845278 5824 11 1562181 0 8273 0 0 10827011 5928 24 1605532 0 8159 0 0 10697129 5718 7 1665896 0 7858 0 0 10173396 5852 9 2098723 0 8145 0 0 10574881 5885 59 1812857 0 8120 0 0 10495941 5784 17 1873823 0 8287 0 0 10955045 5746 41 1437687 0 input (vlan2) output packets errs idrops bytes packets errs bytes colls 8179 0 0 10604523 5992 78 1772952 0 8114 0 0 10595869 5861 7 1794099 0 8230 0 0 10644911 5928 65 1745887 0 8252 0 0 10825497 5898 51 1576670 0 8277 0 0 10739635 5946 20 1660457 0 8375 0 0 10763467 6051 10 1643478 0 8203 0 0 10712495 5921 27 1682199 0 wget: [root@dev /home/emz]# wget http://files2.enaza.ru/isk.7z --2013-05-29 18:03:37-- http://files2.enaza.ru/isk.7z Resolving files2.enaza.ru... 212.33.230.17 Connecting to files2.enaza.ru|212.33.230.17|:80... connection established. Request HTTP sent, waiting for answer... 200 OK Length: 313227902 (299M) [application/x-7z-compressed] Saving to: <> 40% [==============================> ] 125 558 720 6,34M/s ост 30s So... It's obvious that this flow doesn't match any of the configured queues. In this case I expect it to match the default queue. It's clear that the rate in the default queue is far below 60Mbit/sec. Does this output means that this flow isn't shaped by the default queue ? Or may be it means that pfctl -vvvs queue doesn't _show_ the packets that go in the default queue, except on the ones that are explicitely bound to it (I have such packets, I explicitely sent some traffic to the deafult queue) ? Thanks. Eugene.