Ok, here it is:
This is the QoS policy I have on my BGP routers.
It is a fairly basic policy that allows me to dynamically adjust
bandwidth based on need and priority.
You will note that this QoS policy is divided up into three policies or
channels.
Channel #1 is the highest priority and this is where all of my ssh
session traffic and windows terminal server traffic goes.
Essentially this is RDP and NX traffic as all of the my desktops in the
office, and also on our forklifts are a combination of KDE and Windows
using NX clients.
NX works over port 22 and so on.
Notice that the recipe has three channels and these channels have to add
up to and not exceed the UPRATE.
(i.e. PRIORATE1+PRIORATE2+PRIORATE3=UPRATE)
Right now for example my upstream capacity on my cable line is 1Mbit.
So, in order for queueing to happen so I can properly prioritize the
traffic, I have to reduce the upstream rate to a point where it is
backing up on the BGP router, not the cable modem. I found 768K to be
very reliable in all cases.
You have to get that right otherwise QoS may be erratic or simply not work.
I think the script is pretty self documenting but if you want me to
spell out the details about how it works, and what is what with the
iptables portion of the script I would be happy to answer any questions.
-gc
#! /bin/sh
##
# Author: Gregory Carter <gcarter@xxxxxxxxx>
#
# /etc/init.d/qos
#
### BEGIN INIT INFO
# Provides: qos
# Required-Start: $network
# Required-Stop:
# Default-Start: 2 3 5
# Default-Stop:
# Description: quality of service policy gurantees for network services
### END INIT INFO
# Shell functions sourced from /etc/rc.status:
# rc_check check and set local and overall rc status
# rc_status check and set local and overall rc status
# rc_status -v ditto but be verbose in local rc status
# rc_status -v -r ditto and clear the local rc status
# rc_failed set local and overall rc status to failed
# rc_failed <num> set local and overall rc status to <num><num>
# rc_reset clear local rc status (overall remains)
# rc_exit exit appropriate to overall rc status
. /etc/rc.status
# First reset status of this service
rc_reset
# Return values acc. to LSB for all commands but status:
# 0 - success
# 1 - generic or unspecified error
# 2 - invalid or excess argument(s)
# 3 - unimplemented feature (e.ga "reload")
# 4 - insufficient privilege
# 5 - program is not installed
# 6 - program is not configured
# 7 - program is not running
#
# Note that starting an already running service, stopping
# or restarting a not-running service as well as the restart
# with force-reload (in case signalling is not supported) are
# considered a success.
# Real Physical Interface
IFACE=eth2
# Priority Channels
MARKPRIO1="1"
MARKPRIO2="2"
MARKPRIO3="3"
# Master Rate
UPRATE="768"
# Channel Rates
PRIORATE1="340"
PRIORATE2="128"
PRIORATE3="300"
case "$1" in
start)
echo -n "Starting Quality of Service Policy"
## Start daemon with startproc(8). If this fails
## the echo return value is set appropriate.
echo "TOS Settings"
# Set priorities by marking packets.
#
# TOS Matching
#
# Make sure the TOS headers in each TCP transaction are prioritized
first
#
iptables -t mangle -A PREROUTING -m tos --tos
Minimize-Delay -j MARK --set-mark $MARKPRIO1
iptables -t mangle -A PREROUTING -m tos --tos
Minimize-Delay -j RETURN
iptables -t mangle -A PREROUTING -m tos --tos
Minimize-Cost -j MARK --set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -m tos --tos
Minimize-Cost -j RETURN
iptables -t mangle -A PREROUTING -m tos --tos
Maximize-Throughput -j MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -m tos --tos
Maximize-Throughput -j RETURN
iptables -t mangle -A OUTPUT -m tos --tos Minimize-Delay
-j MARK --set-mark $MARKPRIO1
iptables -t mangle -A OUTPUT -m tos --tos Minimize-Delay
-j RETURN
iptables -t mangle -A OUTPUT -m tos --tos Minimize-Cost
-j MARK --set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -m tos --tos Minimize-Cost
-j RETURN
iptables -t mangle -A OUTPUT -m tos --tos Maximize-Throughput
-j MARK --set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -m tos --tos Maximize-Throughput
-j RETURN
# Maximum Priority Services
echo "SSH Priority 1"
# ssh
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 22 -j
MARK --set-mark $MARKPRIO1
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 22 -j
RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 22 -j
MARK --set-mark $MARKPRIO1
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 22 -j
RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 22 -j MARK
--set-mark $MARKPRIO1
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 22 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 22 -j MARK
--set-mark $MARKPRIO1
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 22 -j RETURN
echo "Terminal Services Priority 1"
# Terminal Services
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 3389 -j
MARK --set-mark $MARKPRIO1
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 3389 -j
RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 3389 -j
MARK --set-mark $MARKPRIO1
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 3389 -j
RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 3389 -j
MARK --set-mark $MARKPRIO1
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 3389 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 3389 -j
MARK --set-mark $MARKPRIO1
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 3389 -j RETURN
# Services trying to establish an initial connection between
locations must be
# at highest priority as they are already slow over the back haul.
echo "TCP Connection States Priority 1"
iptables -t mangle -I PREROUTING -p tcp -m tcp --tcp-flags
SYN,RST,ACK SYN -j MARK --set-mark $MARKPRIO1
iptables -t mangle -I PREROUTING -p tcp -m tcp --tcp-flags
SYN,RST,ACK SYN -j RETURN
iptables -t mangle -I OUTPUT -p tcp -m tcp --tcp-flags
SYN,RST,ACK SYN -j MARK --set-mark $MARKPRIO1
iptables -t mangle -I OUTPUT -p tcp -m tcp --tcp-flags
SYN,RST,ACK SYN -j RETURN
# Prio 2
#
# 2nd Class Priority Services
#
echo "Freightlink Clients Priority 2"
# Freightlink Client
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 52272
-j MARK --set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 52272
-j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 52272
-j MARK --set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 52272
-j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 52272 -j
MARK --set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 52272 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 52272 -j
MARK --set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 52272 -j RETURN
echo "DNS Priority 2"
# DNS
iptables -t mangle -A PREROUTING -p udp -m udp --sport 53 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p udp -m udp --sport 53 -j RETURN
iptables -t mangle -A PREROUTING -p udp -m udp --dport 53 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p udp -m udp --dport 53 -j RETURN
iptables -t mangle -A OUTPUT -p udp -m udp --sport 53 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p udp -m udp --sport 53 -j RETURN
iptables -t mangle -A OUTPUT -p udp -m udp --dport 53 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p udp -m udp --dport 53 -j RETURN
echo "SMTP Priority 2"
# SMTP
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 25 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 25 -j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 25 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 25 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 25 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 25 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 25 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 25 -j RETURN
echo "POP3 Priority 2"
# POP3
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 110 -j
MARK --set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 110 -j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 110 -j
MARK --set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 110 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 110 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 110 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 110 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 110 -j RETURN
echo "LDAP Priority 2"
# LDAP
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 389 -j
MARK --set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 389 -j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 389 -j
MARK --set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 389 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 389 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 389 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 389 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 389 -j RETURN
# LDAPS
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 636 -j
MARK --set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 636 -j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 636 -j
MARK --set-mark $MARKPRIO2
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 636 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 636 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 636 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 636 -j MARK
--set-mark $MARKPRIO2
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 636 -j RETURN
# Prio 3
#
# 3th Class Priority Services
#
echo "SMB over TCP Priority 3 (Bulk Default Channel)"
# Microsoft SMB over TCP File Server Traffic
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 445 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 445 -j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 445 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 445 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 445 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 445 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 445 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 445 -j RETURN
echo "SMB over NETBIOS Priority 3 (Bulk Default Channel)"
# Microsoft SMB over NETBIOS Name Service
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 137 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 137 -j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 137 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 137 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 137 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 137 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 137 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 137 -j RETURN
echo "SMB over NETBIOS Datagram Service Priority 3 (Bulk Default
Channel)"
# Microsoft SMB over NETBIOS Datagram Service
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 138 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 138 -j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 138 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 138 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 138 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 138 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 138 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 138 -j RETURN
echo "SMB over NETBIOS Session Service Priority 3 (Bulk Default
Channel)"
# Microsoft SMB over NETBIOS Session Service
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 139 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 139 -j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 139 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 139 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 139 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 139 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 139 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 139 -j RETURN
echo "AXIS Web Services Gateway Priority 3 (Bulk Default Channel)"
# AXIS Web Services Gateway Traffic
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 8080 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --sport 8080 -j RETURN
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 8080 -j
MARK --set-mark $MARKPRIO3
iptables -t mangle -A PREROUTING -p tcp -m tcp --dport 8080 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 8080 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --sport 8080 -j RETURN
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 8080 -j MARK
--set-mark $MARKPRIO3
iptables -t mangle -A OUTPUT -p tcp -m tcp --dport 8080 -j RETURN
echo "Set Interface Target Q"
# Set queue length for IFACE
ifconfig $IFACE txqueuelen 16
echo "Setting Queue Type...."
# Specify queue discipline
tc qdisc add dev $IFACE root handle 1:0 htb default 103 r2q 1
echo "Setting Root Class"
# Set root class
tc class add dev $IFACE parent 1:0 classid 1:1 htb rate
${UPRATE}kbit ceil ${UPRATE}kbit
# Specify sub classes
tc class add dev $IFACE parent 1:1 classid 1:101 htb rate
${PRIORATE1}kbit ceil ${UPRATE}kbit prio 0
tc class add dev $IFACE parent 1:1 classid 1:102 htb rate
${PRIORATE2}kbit ceil ${UPRATE}kbit prio 1
tc class add dev $IFACE parent 1:1 classid 1:103 htb rate
${PRIORATE3}kbit ceil ${UPRATE}kbit prio 2
# Filter packets
tc filter add dev $IFACE parent 1:0 protocol ip prio 1 handle
$MARKPRIO1 fw classid 1:101
tc filter add dev $IFACE parent 1:0 protocol ip prio 1 handle
$MARKPRIO2 fw classid 1:102
tc filter add dev $IFACE parent 1:0 protocol ip prio 1 handle
$MARKPRIO3 fw classid 1:103
# Add queuing disciplines
tc qdisc add dev $IFACE parent 1:101 sfq perturb 16
tc qdisc add dev $IFACE parent 1:102 sfq perturb 16
tc qdisc add dev $IFACE parent 1:103 sfq perturb 16
# Remember status and be verbose
rc_status -v
;;
stop)
echo -n "Removing QoS Policies.....done"
## Stop daemon with killproc(8) and if this fails
## set echo the echo return value.
tc qdisc del dev $IFACE root handle 1:0 htb default 103 r2q 1
rc_status -v
;;
status)
echo -n "Currrent QoS queue stats: "
## Check status with checkproc(8), if process is running
## checkproc will return with exit status 0.
# Status has a slightly different for the status command:
# 0 - service running
# 1 - service dead, but /var/run/ pid file exists
# 2 - service dead, but /var/lock/ lock file exists
# 3 - service not running
# NOTE: checkproc returns LSB compliant status values.
tc -s -d class show dev $IFACE
rc_status -v
;;
*)
echo "Usage: $0
{start|stop|status|try-restart|restart|force-reload|reload}"
exit 1
;;
esac
rc_exit
Scott van Looy wrote:
ooh, please!
On Feb 18 Gregory Carter did spake thusly:
I can post a simple tc / iptables recipe if need be to address this
problem.
-gc
Martijn Lievaart wrote:
Scott van Looy wrote:
Today Martijn Lievaart did spake thusly:
Scott van Looy wrote:
I'm on DSL, I only have small upstream and big downstream.
Everything goes through a 1:1 NAT iptables firewall - so internal
IPs are directly mapped to an external IP. I want to throttle the
internal upstream bandwidth so the internal machines can't upload
at more than 25k/s as the maxing out of the upstream is killing
the much faster downstream...is there a way to do this using
iptables?
Google for wondershaper.
I use wondershaper, it doesn't let me throttle on a per machine
basis tho - I want the firewall to only allow a max of 25k out for
each of my internal machines
Ah, then you have to do some work yourself. The answer is most
probably tc, not netfilter. If you don't have a lot of internal
machines, you can add a filter/qdisc for all of them but for large
amounts of internal machines (or with dhcp), this gets unwieldy.
However, wondershaper should do what you ultimately want, shape
traffic so uploads don't disturb downloads. It works for me.
M4
