Traffic Shaping/Control Tom Eastep Arne Bernin 2001-2008 Thomas M. Eastep 2005 Arne Bernin & Thomas M. Eastep Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, with no Front-Cover, and with no Back-Cover Texts. A copy of the license is included in the section entitled GNU Free Documentation License. Traffic shaping is complex and the Shorewall community is not well equipped to answer traffic shaping questions. So if you are the type of person who needs "insert tab A into slot B" instructions for everything that you do, then please don't try to implement traffic shaping using Shorewall. You will just frustrate yourself and we won't be able to help you. Said another way, reading just Shorewall documentation is not going to give you enough background to use this material. At a minimum, you will need to refer to at least the following additional information: The LARTC HOWTO: http://www.lartc.org The HTB User's Guide: http://luxik.cdi.cz/~devik/qos/htb/manual/userg.htm Some of the documents listed at http://www.netfilter.org/documentation/index.html#documentation-howto. The tutorial by Oskar Andreasson is particularly good. The output of man iptables
Introduction Starting with Version 2.5.5, Shorewall has builtin support for traffic shaping and control. Before this version, the support was quite limited. You were able to use your own tcstart script (and you still are), but besides the tcrules file it was not possible to define classes or queuing disciplines inside the Shorewall config files. The support for traffic shaping and control still does not cover all options available (and especially all algorithms that can be used to queue traffic) in the Linux kernel but it should fit most needs. If you are using your own script for traffic control and you still want to use it in the future, you will find information on how to do this, later in this document. But for this to work, you will also need to enable traffic shaping in the kernel and Shorewall as covered by the next sections.
Linux traffic shaping and control This section gives a brief introduction of how controlling traffic with the Linux kernel works. Although this might be enough for configuring it in the Shorewall configuration files, we strongly recommend that you take a deeper look into the Linux Advanced Routing and Shaping HOWTO. At the time of writing this, the current version is 1.0.0. Since kernel 2.2 Linux has extensive support for controlling traffic. You can define different algorithms that are used to queue the traffic before it leaves an interface. The standard one is called pfifo and is (as the name suggests) of the type First In First out. This means, that it does not shape anything, if you have a connection that eats up all your bandwidth, this queuing algorithm will not stop it from doing so. For Shorewall traffic shaping we use two algorithms, one is called HTB (Hierarchical Token Bucket) and SFQ (Stochastic Fairness Queuing). SFQ is easy to explain: it just tries to track your connections (tcp or udp streams) and balances the traffic between them. This normally works well. HTB allows you to define a set of classes, and you can put the traffic you want into these classes. You can define minimum and maximum bandwidth settings for those classes and order them hierarchically (the less prioritized classes only get bandwidth if the more important have what they need). Shorewall builtin traffic shaping allows you to define these classes (and their bandwidth limits), and it uses SFQ inside these classes to make sure, that different data streams are handled equally. If you are running Shorewall-shell or if you are running Shorewall-perl 4.1.5 or earlier:
You can only shape outgoing traffic. The reason for this is simple, the packets were already received by your network card before you can decide what to do with them. So the only choice would be to drop them which normally makes no sense (since you received the packet already, it went through the possible bottleneck (the incoming connection). The next possible bottleneck might come if the packet leaves on another interface, so this will be the place where queuing might occur. So, defining queues for incoming packets is not very useful, you just want to have it forwarded to the outgoing interface as fast as possible. There is one exception, though. Limiting incoming traffic to a value a bit slower than your actual line speed will avoid queuing on the other end of that connection. This is mostly useful if you don't have access to traffic control on the other side and if this other side has a faster network connection than you do (the line speed between the systems is the bottleneck, e.g. a DSL or Cable Modem connection to your provider's router, the router itself is normally connected to a much faster backbone). So, if you drop packets that are coming in too fast, the underlying protocol might recognize this and slow down the connection. TCP has a builtin mechanism for this, UDP has not (but the protocol over UDP might recognize it , if there is any). The reason why queuing is bad in these cases is, that you might have packets which need to be prioritized over others, e.g. VoIP or ssh. For this type of connections it is important that packets arrive in a certain amount of time. For others like HTTP downloads, it does not really matter if it takes a few seconds more. If you have a large queue on the other side and the router there does not care about QoS or the QoS bits are not set properly, your important packets will go into the same queue as your less time critical download packets which will result in a large delay.
If you are running Shorewall-perl 4.1.6 or later:
You can shape incoming traffic through use of an Intermediate Frame Block (IFB) device. See below. But beware: using an IFB can result in queues building up both at your ISPs router and at your own.
This is not to say that you cannot shape download traffic, regardless of which Shorewall release you are running.
If you wish to shape downloads, you can always configure traffic shaping on your firewall's local interface. An example appears below. Again, however, this can result in queues building up both at your ISPs router and at your own.
You shape and control outgoing traffic by assigning the traffic to classes. Each class is associated with exactly one network interface and has a number of attributes: PRIORITY - Used to give preference to one class over another when selecting a packet to send. The priority is a numeric value with 1 being the highest priority, 2 being the next highest, and so on. RATE - The minimum bandwidth this class should get, when the traffic load rises. Classes with a higher priority (lower PRIORITY value) are served even if there are others that have a guaranteed bandwidth but have a lower priority (higher PRIORITY value). CEIL - The maximum bandwidth the class is allowed to use when the link is idle. MARK - Netfilter has a facility for marking packets. Packet marks have a numeric value which is limited in Shorewall to the values 1-255. You assign packet marks to different types of traffic using entries in the /etc/shorewall/tcrules file. One class for each interface must be designated as the default class. This is the class to which unmarked traffic (packets to which you have not assigned a mark value in /etc/shorewall/tcrules) is assigned. Netfilter also supports a mark value on each connection. You can assign connection mark values in /etc/shorewall/tcrules, you can copy the current packet's mark to the connection mark (SAVE), or you can copy the connection mark value to the current packet's mark (RESTORE). For more information, see this article.
Linux Kernel Configuration You will need at least kernel 2.4.18 for this to work, please take a look at the following screenshot for what settings you need to enable. For builtin support, you need the HTB scheduler, the Ingress scheduler, the PRIO pseudoscheduler and SFQ queue. The other scheduler or queue algorithms are not needed. This screen shot shows how I configured QoS in a 2.6.16 Kernel: And here's my recommendation for a 2.6.21 kernel:
Enable TC support in Shorewall You need this support whether you use the builtin support or whether you provide your own tcstart script. To enable the builtin traffic shaping and control in Shorewall, you have to do the following: Set TC_ENABLED to "Internal" in /etc/shorewall/shorewall.conf. Setting TC_ENABLED=Yes causes Shorewall to look for an external tcstart file (See a later section for details). Setting CLEAR_TC parameter in /etc/shorewall/shorewall.conf to Yes will clear the traffic shaping configuration during Shorewall [re]start and Shorewall stop. This is normally what you want when using the builtin support (and also if you use your own tcstart script) The other steps that follow depend on whether you use your own script or the builtin solution. They will be explained in the following sections.
Using builtin traffic shaping/control Shorewall's builtin traffic shaping feature provides a thin layer on top of the ingress qdesc, HTB and SFQ. That translation layer allows you to: Define HTB classes using Shorewall-style column-oriented configuration files. Integrate the reloading of your traffic shaping configuration with the reloading of your packet-filtering and marking configuration. Assign traffic to HTB classes by TOS value. Assign outgoing TCP ACK packets to an HTB class. Assign traffic to HTB classes based on packet mark value. Shorewall's builtin traffic shaping feature is limited to ten (10) devices. Those few features are really all that builtin traffic shaping/control provides; consequently, you need to understand HTB and Linux traffic shaping as well as Netfilter packet marking in order to use the facility. Again, please see the links at top of this article. For defining bandwidths (for either devices or classes) please use kbit or kbps (for Kilobytes per second) and make sure there is NO space between the number and the unit (it is 100kbit not 100 kbit). Using mbit, mbps or a raw number (which means bytes) could be used, but note that only integer numbers are supported (0.5 is not valid). To properly configure the settings for your devices you need to find out the real up- and downstream rates you have. This is especially the case, if you are using a DSL connection or one of another type that do not have a guaranteed bandwidth. Don't trust the values your provider tells you for this; especially measuring the real download speed is important! There are several online tools that help you find out; search for "dsl speed test" on google (For Germany you can use arcor speed check). Be sure to choose a test located near you.
/etc/shorewall/tcdevices This file allows you to define the incoming and outgoing bandwidth for the devices you want traffic shaping to be enabled. That means, if you want to use traffic shaping for a device, you have to define it here. Columns in the file are as follows: INTERFACE - Name of interface. Each interface may be listed only once in this file. You may NOT specify the name of an alias (e.g., eth0:0) here; see FAQ #18. You man NOT specify wildcards here, e.g. if you have multiple ppp interfaces, you need to put them all in here! With Shorewall versions prior to 3.0.8 and 3.2.0 Beta 8, the device named in this column must exist at the time that Shorewall is started, restarted or refreshed. Beginning with Shorewall 3.0.8 and 3.2.0 Beta 8, Shorewall will determine if the device exists and will only configure the device if it does exist. If it doesn't exist, the following warning is issued: WARNING: Device <device name> not found -- traffic-shaping configuration skipped Shorewall assigns a sequential interface number to each interface (the first entry in /etc/shorewall/tcdevices is interface 1, the second is interface 2 and so on) Beginning with Shorewall-perl 4.1.6, you can explicitly specify the interface number by prefixing the interface name with the number and a colon (":"). Example: 1:eth0. IN-BANDWIDTH - The incoming Bandwidth of that interface. Please note that when you use this column, you are not traffic shaping incoming traffic, as the traffic is already received before you could do so. This Column allows you to define the maximum traffic allowed for this interface in total, if the rate is exceeded, the excess packets are dropped. You want this mainly if you have a DSL or Cable Connection to avoid queuing at your providers side. If you don't want any traffic to be dropped set this to a value faster than your interface maximum rate (or to 0 (zero), if you are running Shorewall 3.2.6 or later). To determine the optimum value for this setting, we recommend that you start by setting it significantly below your measured download bandwidth (20% or so). While downloading, measure the ping response time from the firewall to the upstream router as you gradually increase the setting.The optimal setting is at the point beyond which the ping time increases sharply as you increase the setting. OUT-BANDWIDTH - Specify the outgoing bandwidth of that interface. This is the maximum speed your connection can handle. It is also the speed you can refer as "full" if you define the tc classes. Outgoing traffic above this rate will be dropped. OPTIONS (Added in Shorewall-perl 4.1.4) — A comma-separated list of options from the following list: classify If specified, classification of traffic into the various classes is done by CLASSIFY entries in /etc/shorewall/tcrules or by entries in /etc/shorewall/tcfilters. No MARK value will be associated with classes on this interface. REDIRECTED INTERFACES (Added in Shorewall-perl 4.1.6) — Entries are appropriate in this column only if the device in the INTERFACE column names a Intermediate Frame Block (IFB). It lists the physical interfaces that will have their input shaped using classes defined on the IFB. Neither the IFB nor any of the interfaces listed in this column may have an IN-BANDWIDTH specified. You may specify zero (0) or a dash ("-:) in the IN-BANDWIDTH column. IFB devices automatically get the classify option. Suppose you are using PPP over Ethernet (DSL) and ppp0 is the interface for this. The device has an outgoing bandwidth of 500kbit and an incoming bandwidth of 6000kbit #INTERFACE IN-BANDWITH OUT-BANDWIDTH ppp0 6000kbit 500kbit
/etc/shorewall/tcclasses This file allows you to define the actual classes that are used to split the outgoing traffic. INTERFACE - Name of interface. Users of Shorewall-perl 4.1.6 or later may also specify the interface number. Must match the name (or number) of an interface with an entry in /etc/shorewall/tcdevices. If the interface has the classify option in /etc/shorewall/tcdevices, then the interface name or number must be followed by a colon and a class number. Examples: eth0:1, 4:9. Class numbers must be unique for a given interface. MARK - The mark value which is an integer in the range 1-255. You define these marks in the tcrules file, marking the traffic you want to go into the queuing classes defined in here. You can use the same marks for different Interfaces. You must specify "-' in this column if the device specified in the INTERFACE column has the classify option in /etc/shorewall/tcdevices. RATE - The minimum bandwidth this class should get, when the traffic load rises. Please note that first the classes which equal or a lesser priority value are served even if there are others that have a guaranteed bandwidth but a lower priority. If the sum of the RATEs for all classes assigned to an INTERFACE exceed that interfaces's OUT-BANDWIDTH, then the OUT-BANDWIDTH limit will not be honored. CEIL - The maximum bandwidth this class is allowed to use when the link is idle. Useful if you have traffic which can get full speed when more important services (e.g. interactive like ssh) are not used. You can use the value "full" in here for setting the maximum bandwidth to the defined output bandwidth of that interface. PRIORITY - you have to define a priority for the class. packets in a class with a higher priority (=lesser value) are handled before less prioritized ones. You can just define the mark value here also, if you are increasing the mark values with lesser priority. OPTIONS - A comma-separated list of options including the following: default - this is the default class for that interface where all traffic should go, that is not classified otherwise. defining default for exactly one class per interface is mandatory! tos-<tosname> - this lets you define a filter for the given <tosname> which lets you define a value of the Type Of Service bits in the ip package which causes the package to go in this class. Please note, that this filter overrides all mark settings, so if you define a tos filter for a class all traffic having that mark will go in it regardless of the mark on the package. You can use the following for this option: tos-minimize-delay (16) tos-maximize-throughput (8) tos-maximize-reliability (4) tos-minimize-cost (2) tos-normal-service (0) Each of this options is only valid for one class per interface. tcp-ack - if defined causes an tc filter to be created that puts all tcp ack packets on that interface that have an size of <=64 Bytes to go in this class. This is useful for speeding up downloads. Please note that the size of the ack packets is limited to 64 bytes as some applications (p2p for example) use to make every package an ack package which would cause them all into here. We want only packets WITHOUT payload to match, so the size limit. Bigger packets just take their normal way into the classes. This option is only valid for class per interface.
/etc/shorewall/tcrules Unlike rules in the shorewall-rules(5) file, evaluation of rules in this file will continue after a match. So the final mark for each packet will be the one assigned by the LAST tcrule that matches. Also unlike rules in the shorewall-rules(5) file, the tcrules file is not stateful. So every packet that goes into, out of or through your firewall is subject to entries in the tcrules file. Because tcrules are not stateful, it is necessary to understand basic IP socket operation. Here is an edited excerpt from a post on the Shorewall Users list:
For the purposes of this discussion, the world is separated into clients and servers. Servers provide services to clients. When a server starts, it creates a socket and binds the socket to an address. For AF_INET (IPv4) and AF_INET6 (IPv6) sockets, that address is an ordered triple consisting of an IPv4 or IPv6 address, a protocol, and possibly a port number. Port numbers are only used when the protocol is TCP, UDP, SCTP or DCCP. The protocol and port number used by a server are typically well-known so that clients will be able to connect to it or send datagrams to it. So SSH servers bind to TCP port 22, SMTP servers bind to TCP port 25, etc. We will call this port the SERVER PORT. When a client want to use the service provided by a server, it also creates a socket and, like the server's socket, the client's socket must be bound to an address. But in the case of the client, the socket is usually given an automatic address binding. For AF_INET and AF_INET6 sockets. the IP address is the IP address of the client system (loose generalization) and the port number is selected from a local port range. On Linux systems, the local port range can be seen by cat /proc/sys/net/ipv4/ip_local_port_range. So it is not possible in advance to determine what port the client will be using. Whatever it is, we'll call it the CLIENT PORT. Now:
Packets sent from the client to the server will have:
SOURCE PORT = CLIENT PORT DEST PORT = SERVER PORT
Packets sent from the server to the client will have:
SOURCE PORT = SERVER PORT DEST PORT = CLIENT PORT
Since the SERVER PORT is generally the only port known ahead of time, we must categorize traffic from the server to the client using the SOURCE PORT.
The fwmark classifier provides a convenient way to classify packets for traffic shaping. The /etc/shorewall/tcrules file is used for specifying these marks in a tabular fashion. For an in-depth look at the packet marking facility in Netfilter/Shorewall, please see this article. Normally, packet marking occurs in the PREROUTING chain before any address rewriting takes place. This makes it impossible to mark inbound packets based on their destination address when SNAT or Masquerading are being used. You can cause packet marking to occur in the FORWARD chain by using the MARK_IN_FORWARD_CHAIN option in shorewall.conf. Columns in the file are as follows: MARK or CLASSIFY - MARK specifies the mark value is to be assigned in case of a match. This is an integer in the range 1-255. This value may be optionally followed by : and either F, P or "T" to designate that the marking will occur in the FORWARD, PREROUTING or POSTROUTING chains respectively. If this additional specification is omitted, the chain used to mark packets will be determined as follows: If the SOURCE is $FW[:<address>], then the rule is inserted in the OUTPUT chain. Otherwise, the chain is determined by the setting of the MARK_IN_FORWARD_CHAIN option in shorewall.conf. The "T" qualifier was added in Shorewall version 3.3.6 and is not available in earlier versions. Use this qualifier if you want the rule to apply equally to traffic being routed through the firewall and to traffic originating on the firewall itself. Normally, the mark is applied to the packet. If you follow the mark value with ":" and "C", then the mark is applied to the connection. "C" can be combined with "F", "P" or "T" to designate that the connection should be marked in a particular chain (e.g., "CF", "CP", "CT"). There are additional special values available: RESTORE[/mask] -- restore the packet's mark from the connection's mark using the supplied mask if any. Your kernel and iptables must include CONNMARK support. As above, may be followed by :P, :F or :T. SAVE[/mask] -- save the packet's mark to the connection's mark using the supplied mask if any. Your kernel and iptables must include CONNMARK support. As above, may be followed by :P, :F or :T. CONTINUE Don't process any more marking rules in the table. As above, may be followed by :P, :F or :T. COMMENT (Added in Shorewall version 3.3.3) -- the rest of the line will be attached as a comment to the Netfilter rule(s) generated by the following entries. The comment will appear delimited by "/* ... */" in the output of shorewall show mangle To stop the comment from being attached to further rules, simply include COMMENT on a line by itself. To use CLASSIFY, your kernel and iptables must include CLASSIFY target support. In that case, this column contains a classification (classid) of the form <major>:<minor> where <major> and <minor> are integers. Corresponds to the 'class' specification in these traffic shaping modules: atm cbq dsmark pfifo_fast htb prio With Shorewall versions prior to 3.2.3, classify rules are always placed in the POSTROUTING chain. Beginning with Shorewall 3.2.3, classification occurs in the POSTROUTING chain except when the SOURCE contains $FW[:<address>] in which case, the classify action takes place in the OUTPUT chain. When used with the builtin traffic shaper, the <major> class is the interface number and the <minor> class is either a) the MARK value of the class preceded by the number "1" (MARK value 1 is <minor> class 11, MARK value 22 is <minor> class 122, and so on) or b) The class number (if the classify option was specified in for the interface /etc/shorewall/interfaces) SOURCE - Source of the packet. May be: An interface name - matches traffic entering the firewall on the specified interface. May not be used in classify rules or in rules using the :T chain qualifier. A comma-separated list of host or network IP addresses or MAC addresses. This form will not match traffic that originates on the firewall itself unless either <major><minor> or the :T chain qualifier is used in the MARK column. Examples: 0.0.0.0/0 192.168.1.0/24, 172.20.4.0/24 An interface name followed by a colon (":") followed by a comma-separated list of host or network IP addresses or MAC addresses. May not be used in classify rules or in rules using the :T chain qualifier. $FW optionally followed by a colon (":") and a comma-separated list of host or network IP addresses. matches packets originating on the firewall. May not be used with a chain qualifier (:P, :F, etc.) in the MARK column. MAC addresses must be prefixed with "~" and use "-" as a separator. Example: ~00-A0-C9-15-39-78 If your kernel includes iprange match support, then address ranges may be included in the address lists. DEST - Destination of the packet. May be: An interface name. May not be used in the PREROUTING chain (:P in the mark column or no chain qualifier and MARK_IN_FORWARD_CHAIN=No in shorewall.conf (5)). The interface name may be optionally followed by a colon (":") and an IP address list. A comma-separated list of host or network IP addresses. The list may include ip address ranges if your kernel and iptables include iprange support. PROTO - Protocol - Must be "tcp", "udp", "icmp", "ipp2p", "ipp2p:udp", "ipp2p:all" a number, or "all". "ipp2p" requires ipp2p match support in your kernel and iptables. PORT(S) - Destination Ports. A comma-separated list of Port names (from /etc/services), port numbers or port ranges; if the protocol is "icmp", this column is interpreted as the destination icmp-type(s). If the protocol is ipp2p, this column is interpreted as an ipp2p option without the leading "--" (example "bit" for bit-torrent). If no PORT is given, "ipp2p" is assumed. Note that the xtables-addons version of IPP2P does not support the "ipp2p" option; to use that version of IPP2P with Shorewall-shell or with Shorewall-perl 4.2.4 or earlier, you must specify an option other than "ipp2p". Shorewall-perl 4.2.5 and later support a comma-separated list of IPP2P options in this column; if the column is empty or contains "ipp2p", then those versions of Shorewall-perl will substitute "edk,kazaa,gnu,dc". This column is ignored if PROTOCOL = all but must be entered if any of the following field is supplied. In that case, it is suggested that this field contain "-" CLIENT PORT(S) - (Optional) Port(s) used by the client. If omitted, any source port is acceptable. Specified as a comma-separate list of port names, port numbers or port ranges. USER/GROUP (Added in Shorewall version 1.4.10) - (Optional) This column may only be non-empty if the SOURCE is the firewall itself. When this column is non-empty, the rule applies only if the program generating the output is running under the effective user and/or group. It may contain : [!][<user name or number>]:[<group name or number>][+<program name>] The colon is optional when specifying only a user. Examples: joe #program must be run by joe :kids #program must be run by a member of the 'kids' group !:kids #program must not be run by a member of the 'kids' group +upnpd #program named upnpd (This feature was removed from Netfilter in kernel version 2.6.14). TEST - Defines a test on the existing packet or connection mark. The rule will match only if the test returns true. Tests have the format [!]<value>[/<mask>][:C] Where: ! Inverts the test (not equal) <value> Value of the packet or connection mark. <mask> A mask to be applied to the mark before testing :C Designates a connection mark. If omitted, the packet mark's value is tested. LENGTH (Optional, added in Shorewall version 3.2.0) Packet Length - This field, if present, allows you to match the length of a packet against a specific value or range of values. A range is specified in the form <min>:<max> where either <min> or <max> (but not both) may be omitted. If <min> is omitted, then 0 is assumed; if <max> is omitted, than any packet that is <min> or longer will match. You must have iptables length support for this to work. If you let it empty or place an "-" here, no length match will be done. Examples: 1024, 64:1500, :100 TOS (Optional, added in Shorewall version 3.2.0 Beta 6) Type of Service. Either a standard name, or a numeric value to match.
Minimize-Delay (16) Maximize-Throughput (8) Maximize-Reliability (4) Minimize-Cost (2) Normal-Service (0)
HELPER (Optional, added in Shorewall version 4.2.0 Beta 2). Names one of the Netfilter protocol helper modules such as ftp, sip, amanda, etc.
All packets arriving on eth1 should be marked with 1. All packets arriving on eth2 and eth3 should be marked with 2. All packets originating on the firewall itself should be marked with 3. #MARK SOURCE DESTINATION PROTOCOL PORT(S) 1 eth1 0.0.0.0/0 all 2 eth2 0.0.0.0/0 all 2 eth3 0.0.0.0/0 all 3 $FW 0.0.0.0/0 all All GRE (protocol 47) packets destined for 155.186.235.151 should be marked with 12. #MARK SOURCE DESTINATION PROTOCOL PORT(S) 12:T 0.0.0.0/0 155.182.235.151 47 All SSH request packets originating in 192.168.1.0/24 and destined for 155.186.235.151 should be marked with 22. #MARK SOURCE DESTINATION PROTOCOL PORT(S) 22:T 192.168.1.0/24 155.182.235.151 tcp 22 All SSH packets packets going out of the first device in in /etc/shorewall/tcdevices should be assigned to the class with mark value 10. #MARK SOURCE DESTINATION PROTOCOL PORT(S) CLIENT # PORT(S) 1:110 0.0.0.0/0 0.0.0.0/0 tcp 22 1:110 0.0.0.0/0 0.0.0.0/0 tcp - 22 Mark all ICMP echo traffic with packet mark 1. Mark all peer to peer traffic with packet mark 4. This is a little more complex than otherwise expected. Since the ipp2p module is unable to determine all packets in a connection are P2P packets, we mark the entire connection as P2P if any of the packets are determined to match. We assume packet/connection mark 0 to means unclassified. Traffic originating on the firewall is not covered by this example. #MARK SOURCE DESTINATION PROTOCOL PORT(S) CLIENT USER/ TEST # PORT(S) GROUP 1 0.0.0.0/0 0.0.0.0/0 icmp echo-request 1 0.0.0.0/0 0.0.0.0/0 icmp echo-reply RESTORE 0.0.0.0/0 0.0.0.0/0 all - - - 0 CONTINUE 0.0.0.0/0 0.0.0.0/0 all - - - !0 4 0.0.0.0/0 0.0.0.0/0 ipp2p:all SAVE 0.0.0.0/0 0.0.0.0/0 all - - - !0 The last four rules can be translated as:
"If a packet hasn't been classified (packet mark is 0), copy the connection mark to the packet mark. If the packet mark is set, we're done. If the packet is P2P, set the packet mark to 4. If the packet mark has been set, save it to the connection mark."
Mark all forwarded VOIP connections with connection mark 1 and ensure that all VOIP packets also receive that mark (assumes that nf_conntrack_sip is loaded and that Shorewall-perl 4.2.0 or later is being used). #MARK SOURCE DESTINATION PROTOCOL PORT(S) CLIENT USER/ TEST CONNBYTES TOS HELPER # PORT(S) GROUP RESTORE 0.0.0.0/0 0.0.0.0/0 all - - - 0 CONTINUE 0.0.0.0/0 0.0.0.0/0 all - - - !0 1 0.0.0.0/0 0.0.0.0/0 all - - - - - - sip SAVE 0.0.0.0/0 0.0.0.0/0 all - - - !0
ppp devices If you use ppp/pppoe/pppoa) to connect to your Internet provider and you use traffic shaping you need to restart shorewall traffic shaping. The reason for this is, that if the ppp connection gets restarted (and it usually does this at least daily), all tc filters/qdiscs related to that interface are deleted. The easiest way to achieve this, is just to restart shorewall once the link is up. To achieve this add a small executable script to/etc/ppp/ip-up.d. #! /bin/sh /sbin/shorewall refresh
Real life examples
A Shorewall User's Experience Chuck Kollars has provided an excellent writeup about his traffic shaping experiences.
Configuration to replace Wondershaper You are able to fully replace the wondershaper script by using the buitin traffic control.You can find example configuration files at "http://www1.shorewall.net/pub/shorewall/Samples/tc4shorewall/. Please note that they are just examples and need to be adjusted to work for you. In this example it is assumed that your interface for your Internet connection is ppp0 (for DSL), if you use another connection type, you have to change it. You also need to change the settings in the tcdevices.wondershaper file to reflect your line speed. The relevant lines of the config files follow here. Please note that this is just a 1:1 replacement doing exactly what wondershaper should do. You are free to change it...
tcdevices file #INTERFACE IN-BANDWITH OUT-BANDWIDTH ppp0 5000kbit 500kbit
tcclasses file #INTERFACE MARK RATE CEIL PRIORITY OPTIONS ppp0 1 5*full/10 full 1 tcp-ack,tos-minimize-delay ppp0 2 3*full/10 9*full/10 2 default ppp0 3 2*full/10 8*full/10 2
tcrules file #MARK SOURCE DEST PROTO PORT(S) CLIENT USER # PORT(S) 1:F 0.0.0.0/0 0.0.0.0/0 icmp echo-request 1:F 0.0.0.0/0 0.0.0.0/0 icmp echo-reply # mark traffic which should have a lower priority with a 3: # mldonkey 3 0.0.0.0/0 0.0.0.0/0 udp - 4666 Wondershaper allows you to define a set of hosts and/or ports you want to classify as low priority. To achieve this , you have to add these hosts to tcrules and set the mark to 3 (true if you use the example configuration files).
Setting hosts to low priority lets assume the following settings from your old wondershaper script (don't assume these example values are really useful, they are only used for demonstrating ;-): # low priority OUTGOING traffic - you can leave this blank if you want # low priority source netmasks NOPRIOHOSTSRC="192.168.1.128/25 192.168.3.28" # low priority destination netmasks NOPRIOHOSTDST=60.0.0.0/24 # low priority source ports NOPRIOPORTSRC="6662 6663" # low priority destination ports NOPRIOPORTDST="6662 6663" This would result in the following additional settings to the tcrules file: 3 192.168.1.128/25 0.0.0.0/0 all 3 192.168.3.28 0.0.0.0/0 all 3 0.0.0.0/0 60.0.0.0/24 all 3 0.0.0.0/0 0.0.0.0/0 udp 6662,6663 3 0.0.0.0/0 0.0.0.0/0 udp - 6662,6663 3 0.0.0.0/0 0.0.0.0/0 tcp 6662,6663 3 0.0.0.0/0 0.0.0.0/0 tcp - 6662,6663
A simple setup This is a simple setup for people sharing an Internet connection and using different computers for this. It just basically shapes between 2 hosts which have the ip addresses 192.168.2.23 and 192.168.2.42
tcdevices file #INTERFACE IN-BANDWITH OUT-BANDWIDTH ppp0 6000kbit 700kbit We have 6mbit down and 700kbit upstream.
tcclasses file #INTERFACE MARK RATE CEIL PRIORITY OPTIONS ppp0 1 10kbit 50kbit 1 tcp-ack,tos-minimize-delay ppp0 2 300kbit full 2 ppp0 3 300kbit full 2 ppp0 4 90kbit 200kbit 3 default We add a class for tcp ack packets with highest priority, so that downloads are fast. The following 2 classes share most of the bandwidth between the 2 hosts, if the connection is idle, they may use full speed. As the hosts should be treated equally they have the same priority. The last class is for the remaining traffic.
tcrules file #MARK SOURCE DEST PROTO PORT(S) CLIENT USER # PORT(S) 1:F 0.0.0.0/0 0.0.0.0/0 icmp echo-request 1:F 0.0.0.0/0 0.0.0.0/0 icmp echo-reply 2:F 192.168.2.23 0.0.0.0/0 all 3:F 192.168.2.42 0.0.0.0/0 all We mark icmp ping and replies so they will go into the fast interactive class and set a mark for each host.
A Warning to Xen Users If you are running traffic shaping in your dom0 and traffic shaping doesn't seem to be limiting outgoing traffic properly, it may be due to "checksum offloading" in your domU(s). Check the output of "shorewall show tc". Here's an excerpt from the output of that command: class htb 1:130 parent 1:1 leaf 130: prio 3 quantum 1500 rate 76000bit ceil 230000bit burst 1537b/8 mpu 0b overhead 0b cburst 1614b/8 mpu 0b overhead 0b level 0 Sent 559018700 bytes 75324 pkt (dropped 0, overlimits 0 requeues 0) rate 299288bit 3pps backlog 0b 0p requeues 0 lended: 53963 borrowed: 21361 giants: 90174 tokens: -26688 ctokens: -14783 There are two obvious problems in the above output: The rate (299288) is considerably larger than the ceiling (230000). There are a large number (90174) of giants reported. This problem will be corrected by disabling "checksum offloading" in your domU(s) using the ethtool utility. See the one of the Xen articles for instructions.
Shaping Download Traffic As stated at the outset, traffic shaping works on traffic being sent by the firewall. Download traffic from the Internet to local hosts is sent by the firewall over a local interface. So it follows that if you want to shape such traffic, you must configure shaping on the local interface. Shaping of download traffic is most straightforward when there are only two interface. That way, traffic leaving the local interface falls into only two broad categories: Traffic being forwarded from the Internet Traffic that originated on the firewall itself In general, you will want to shape the forwarded traffic and leave the local traffic unrestricted. Extending the simple example above: /etc/shorewall/tcdevices:#INTERFACE IN-BANDWITH OUT-BANDWIDTH ppp0 6000kbit 700kbit eth1 - 100mbit /etc/shorewall/tcclasses:#INTERFACE MARK RATE CEIL PRIORITY OPTIONS ppp0 1 10kbit 50kbit 1 tcp-ack,tos-minimize-delay ppp0 2 300kbit full 2 ppp0 3 300kbit full 2 ppp0 4 90kbit 200kbit 3 default eth0 1 100kbit 500kbit 1 tcp-ack eth0 2 3mbit 6mbit 2 eth0 3 3mbit 6mbit 3 eth0 4 94mbit full default #for local traffic /etc/shorewall/tcrules:#MARK SOURCE DEST PROTO PORT(S) CLIENT USER # PORT(S) 1:F 0.0.0.0/0 0.0.0.0/0 icmp echo-request 1:F 0.0.0.0/0 0.0.0.0/0 icmp echo-reply 2:F 192.168.2.23 0.0.0.0/0 all 3:F 192.168.2.42 0.0.0.0/0 all 2:F ppp0 192.168.2.23 all 3:F ppp0 192.168.2.42 all
Intermediate Frame Block (IFB) Devices Beginning with Shorewall 4.1.6, Shorewall-perl includes support for IFBs. The principles behind an IFB is fairly simple: It looks like a network interface although it is never given an IPv4 configuration. Because it is a network interface, queuing disciplines can be associated with an IFB. The magic of an IFB comes in the fact that a filter may be defined on a real network interface such that each packet that arrives on that interface is queued for the IFB! In that way, the IFB provides a means for shaping input traffic. To use an IFB, you must have IFB support in your kernel (configuration option CONFIG_IFB). Assuming that you have a modular kernel, the name of the IFB module is 'ifb' and may be loaded using the command modprobe ifb (if you have modprobe installed) or insmod /path/to/module/ifb. By default, two IFB devices (ifb0 and ifb1) are created. You can control that using the numifbs option (e.g., modprobe ifb numifbs=1). To create a single IFB when Shorewall starts, place the following two commands in /etc/shorewall/init: modprobe ifb numifbs=1 ip link set ifb0 up Entries in /etc/shorewall/tcrules have no effect on shaping traffic through an IFB. To allow classification of such traffic, the /etc/shorewall/tcfilters file has been added. Entries in that file create u32 classification rules.
/etc/shorewall/tcfilters While this file was created to allow shaping of traffic through an IFB, the file may be used for general traffic classification as well. The file is similar to shorewall-tcrules(5) with the following key exceptions: The first match determines the classification, whereas in the tcrules file, the last match determines the classification. ipsets are not supported DNS Names are not supported filters are applied to packets as they appear on the wire. So incoming packets will not have DNAT applied yet (the destination IP address will be the external address) and outgoing packets will have had SNAT applied. The last point warrants elaboration. When looking at traffic being shaped by an IFB, there are two cases to consider: Requests — packets being sent from remote clients to local servers. These packets may undergo subsequent DNAT, either as a result of entries in /etc/shorewall/nat or as a result of DNAT or REDIRECT rules. Example: /etc/shorewall/rules: #ACTION SOURCE DEST PROTO DEST SOURCE ORIGINAL # PORT(S) PORT(S) DEST DNAT net dmz:192.168.4.5 tcp 80 - 206.124.146.177 Requests redirected by this rule will have destination IP address 206.124.146.177 and destination port 80. Responses — packets being sent from remote servers to local clients. These packets may undergo subsequent DNAT as a result of entries in /etc/shorewall/nat or in /etc/shorewall/masq. The packet's destination IP address will be the external address specified in the entry. Example: /etc/shorewall/masq:#INTERFACE SOURCE ADDRESS eth0 192.168.1.0/24 206.124.146.179 HTTP response packets corresponding to requests that fall under that rule will have destination IP address 206.124.146.179 and source port 80. Columns in the file are as follow. As in all Shorewall configuration files, a hyphen ("-") may be used to indicate that no value is supplied in the column. CLASS The interface name or number followed by a colon (":") and the class number. SOURCE SOURCE IP address (host or network). DNS names are not allowed. DEST DESTINATION IP address (host or network). DNS names are not allowed. PROTO Protocol name or number. DEST PORT(S) Comma-separated list of destination port names or numbers. May only be specified if the protocol is TCP, UDP, SCTP or ICMP. Port ranges are supported except for ICMP. SOURCE PORT Comma-separated list of source port names or numbers. May only be specified if the protocol is TCP, UDP or SCTP. Port ranges are supported. Example: I've used this configuration on my own firewall. The IFB portion is more for test purposes rather than to serve any well-reasoned QOS strategy. /etc/shorewall/init:qt modprobe ifb numifbs=1 qt ip link set dev ifb0 up /etc/shorewall/tcdevices:#INTERFACE IN-BANDWITH OUT-BANDWIDTH OPTIONS REDIRECTED # INTERFACES 1:eth0 - 384kbit classify 2:ifb0 - 1300kbit - eth0 /etc/shorewall/tcclasses:#INTERFACE MARK RATE CEIL PRIORITY OPTIONS 1:110 - 5*full/10 full 1 tcp-ack,tos-minimize-delay 1:120 - 2*full/10 6*full/10 2 default 1:130 - 2*full/10 6*full/10 3 2:110 - 5*full/10 full 1 tcp-ack,tos-minimize-delay 2:120 - 2*full/10 6*full/10 2 default 2:130 - 2*full/10 6*full/10 3/etc/shorewall/tcfilters:#INTERFACE: SOURCE DEST PROTO DEST SOURCE #CLASS PORT(S) PORT(S) # # OUTGOING TRAFFIC # 1:130 206.124.146.178 - tcp - 49441,49442 #BITTORRENT on wookie 1:110 206.124.146.178 #wookie 1:110 206.124.146.179 #SNAT of internal systems 1:110 206.124.146.180 #Work Laptop 1:110 - - icmp echo-request,echo-reply 1:110 - - icmp echo-reply 1:130 206.124.146.177 - tcp - 873,25 #Bulk Traffic # # INCOMING TRAFFIC # 2:110 - 206.124.146.178 #Wookie 2:110 - 206.124.146.179 #SNAT Responses 2:110 - 206.124.146.180 #Work Laptop 2:130 - 206.124.146.177 tcp 25 #Incoming Email. You can examine the installed filters with the shorewall show filters command. What follows shows the output for eth0 with the filters shown above. Bold font are comments explaining the rules.gateway:~ # shorewall-lite show filters Shorewall Lite 4.1.6 Classifiers at gateway - Fri Mar 21 08:06:47 PDT 2008 Device eth1: Device eth2: Device eth0: filter parent 1: protocol ip pref 10 u32 filter parent 1: protocol ip pref 10 u32 fh 3: ht divisor 1 <========= Start of table 3. parses TCP header filter parent 1: protocol ip pref 10 u32 fh 3::800 order 2048 key ht 3 bkt 0 flowid 1:130 (rule hit 102 success 0) match 03690000/ffff0000 at nexthdr+0 (success 0 ) <========= SOURCE PORT 873 goes to class 1:130 filter parent 1: protocol ip pref 10 u32 fh 2: ht divisor 1 <========= Start of table 2. parses ICMP header filter parent 1: protocol ip pref 10 u32 fh 2::800 order 2048 key ht 2 bkt 0 flowid 1:110 (rule hit 0 success 0) match 08000000/ff000000 at nexthdr+0 (success 0 ) <========= ICMP Type 8 goes to class 1:110 filter parent 1: protocol ip pref 10 u32 fh 2::801 order 2049 key ht 2 bkt 0 flowid 1:110 (rule hit 0 success 0) match 00000000/ff000000 at nexthdr+0 (success 0 ) <========= ICMP Type 0 goes to class 1:110 filter parent 1: protocol ip pref 10 u32 fh 1: ht divisor 1 <========= Start of table 1. parses TCP header filter parent 1: protocol ip pref 10 u32 fh 1::800 order 2048 key ht 1 bkt 0 flowid 1:130 (rule hit 0 success 0) match c1210000/ffff0000 at nexthdr+0 (success 0 ) <========= SOURCE PORT 49441 goes to class 1:130 filter parent 1: protocol ip pref 10 u32 fh 1::801 order 2049 key ht 1 bkt 0 flowid 1:130 (rule hit 0 success 0) match c1220000/ffff0000 at nexthdr+0 (success 0 ) <========= SOURCE PORT 49442 goes to class 1:130 filter parent 1: protocol ip pref 10 u32 fh 800: ht divisor 1 <========= Start of Table 800. Packets start here! =============== The following 2 rules are generated by the class definition in /etc/shorewall/classes ================== filter parent 1: protocol ip pref 10 u32 fh 800::800 order 2048 key ht 800 bkt 0 flowid 1:110 (rule hit 2204 success 138) match 00060000/00ff0000 at 8 (success 396 ) <========= TCP match 05000000/0f00ffc0 at 0 (success 250 ) <========= Header length 20 and Packet Length < 64 match 00100000/00ff0000 at 32 (success 138 ) <========= ACK filter parent 1: protocol ip pref 10 u32 fh 800::801 order 2049 key ht 800 bkt 0 flowid 1:110 (rule hit 2066 success 0) match 00100000/00100000 at 0 (success 0 ) <========= Minimize-delay goes to class 1:110 =============== Jump to Table 1 if the matches are met ================== filter parent 1: protocol ip pref 10 u32 fh 800::802 order 2050 key ht 800 bkt 0 link 1: (rule hit 2066 success 0) match ce7c92b2/ffffffff at 12 (success 1039 ) <========= SOURCE 206.124.146.178 match 00060000/00ff0000 at 8 (success 0 ) <========= PROTO TCP offset 0f00>>6 at 0 eat filter parent 1: protocol ip pref 10 u32 fh 800::803 order 2051 key ht 800 bkt 0 flowid 1:110 (rule hit 2066 success 1039) match ce7c92b2/ffffffff at 12 (success 1039 ) <========= SOURCE 206.124.146.178 goes to class 1:110 filter parent 1: protocol ip pref 10 u32 fh 800::804 order 2052 key ht 800 bkt 0 flowid 1:110 (rule hit 1027 success 132) match ce7c92b3/ffffffff at 12 (success 132 ) <========= SOURCE 206.124.146.179 goes to class 1:110 filter parent 1: protocol ip pref 10 u32 fh 800::805 order 2053 key ht 800 bkt 0 flowid 1:110 (rule hit 895 success 603) match ce7c92b4/ffffffff at 12 (success 603 ) <========= SOURCE 206.124.146.180 goes to class 1:110 =============== Jump to Table 2 if the matches are met ================== filter parent 1: protocol ip pref 10 u32 fh 800::806 order 2054 key ht 800 bkt 0 link 2: (rule hit 292 success 0) match 00010000/00ff0000 at 8 (success 0 ) <========= PROTO ICMP offset 0f00>>6 at 0 eat =============== Jump to Table 3 if the matches are met ================== filter parent 1: protocol ip pref 10 u32 fh 800::807 order 2055 key ht 800 bkt 0 link 3: (rule hit 292 success 0) match ce7c92b1/ffffffff at 12 (success 265 ) <========= SOURCE 206.124.146.177 match 00060000/00ff0000 at 8 (success 102 ) <========= PROTO TCP offset 0f00>>6 at 0 eat
Understanding the output of 'shorewall show tc' The shorewall show tc (shorewall-lite show tc) command displays information about the current state of traffic shaping. For each device, it executes the following commands: echo Device $device: tc -s -d qdisc show dev $device echo tc -s -d class show dev $device echo So, the traffic-shaping output is generated entirely by the tc utility. Here's the output of for eth0. The configuration is the one shown in the preceding section (the output was obtained almost 24 hours later than the shorewall show filters output shown above). Device eth0: ============== The primary queuing discipline is HTB (Hierarchical Token Bucket) ==================== qdisc htb 1: r2q 10 default 120 direct_packets_stat 9 ver 3.17 Sent 2133336743 bytes 4484781 pkt (dropped 198, overlimits 4911403 requeues 21) <=========== Note the overlimits and dropped counts rate 0bit 0pps backlog 0b 8p requeues 21 ============== The ingress filter. If you specify IN-BANDWIDTH, you can see the 'dropped' count here. ========= In this case, the packets are being sent to the IFB for shaping qdisc ingress ffff: ---------------- Sent 4069015112 bytes 4997252 pkt (dropped 0, overlimits 0 requeues 0) rate 0bit 0pps backlog 0b 0p requeues 0 ============ Each of the leaf HTB classes has an SFQ qdisc to ensure that each flow gets its turn ============ qdisc sfq 110: parent 1:110 limit 128p quantum 1514b flows 128/1024 perturb 10sec Sent 613372519 bytes 2870225 pkt (dropped 0, overlimits 0 requeues 6) rate 0bit 0pps backlog 0b 0p requeues 6 qdisc sfq 120: parent 1:120 limit 128p quantum 1514b flows 128/1024 perturb 10sec Sent 18434920 bytes 60961 pkt (dropped 0, overlimits 0 requeues 0) rate 0bit 0pps backlog 0b 0p requeues 0 qdisc sfq 130: parent 1:130 limit 128p quantum 1514b flows 128/1024 perturb 10sec Sent 1501528722 bytes 1553586 pkt (dropped 198, overlimits 0 requeues 15) rate 0bit 0pps backlog 11706b 8p requeues 15 ============= Class 1:110 -- the high-priority class =========== Note the rate and ceiling calculated from 'full' class htb 1:110 parent 1:1 leaf 110: prio 1 quantum 4800 rate 192000bit ceil 384000bit burst 1695b/8 mpu 0b overhead 0b cburst 1791b/8 mpu 0b overhead 0b level 0 Sent 613372519 bytes 2870225 pkt (dropped 0, overlimits 0 requeues 0) rate 195672bit 28pps backlog 0b 0p requeues 0 <=========== Note the current rate of traffic. There is no queuing (no packet backlog) lended: 2758458 borrowed: 111773 giants: tokens: 46263 ctokens: 35157 ============== The root class ============ class htb 1:1 root rate 384000bit ceil 384000bit burst 1791b/8 mpu 0b overhead 0b cburst 1791b/8 mpu 0b overhead 0b level 7 Sent 2133276316 bytes 4484785 pkt (dropped 0, overlimits 0 requeues 0) <==== Total output traffic since last 'restart' rate 363240bit 45pps backlog 0b 0p requeues 0 lended: 1081936 borrowed: 0 giants: 0 tokens: -52226 ctokens: -52226 ============= Bulk Class (outgoing rsync, email and bittorrent) ============ class htb 1:130 parent 1:1 leaf 130: prio 3 quantum 1900 rate 76000bit ceil 230000bit burst 1637b/8 mpu 0b overhead 0b cburst 1714b/8 mpu 0b overhead 0b level 0 Sent 1501528722 bytes 1553586 pkt (dropped 198, overlimits 0 requeues 0) rate 162528bit 14pps backlog 0b 8p requeues 0 <======== Queuing is occurring (8 packet backlog). The rate is still below the ceiling. lended: 587134 borrowed: 966459 giants: 0 During peak activity, the rate tops out at around 231000 (just above ceiling). tokens: -30919 ctokens: -97657 ================= Default class (mostly serving web pages) =============== class htb 1:120 parent 1:1 leaf 120: prio 2 quantum 1900 rate 76000bit ceil 230000bit burst 1637b/8 mpu 0b overhead 0b cburst 1714b/8 mpu 0b overhead 0b level 0 Sent 18434920 bytes 60961 pkt (dropped 0, overlimits 0 requeues 0) rate 2240bit 2pps backlog 0b 0p requeues 0 lended: 57257 borrowed: 3704 giants: 0 tokens: 156045 ctokens: 54178
Using your own tc script
Replacing builtin tcstart file If you prefer your own tcstart file, just install it in /etc/shorewall/tcstart. In your tcstart script, when you want to run the tc utility, use the run_tc function supplied by Shorewall if you want tc errors to stop the firewall. Set TC_ENABLED=Yes and CLEAR_TC=Yes Supply an /etc/shorewall/tcstart script to configure your traffic shaping rules. Optionally supply an /etc/shorewall/tcclear script to stop traffic shaping. That is usually unnecessary. If your tcstart script uses the fwmark classifier, you can mark packets using entries in /etc/shorewall/tcrules.
Traffic control outside Shorewall To start traffic shaping when you bring up your network interfaces, you will have to arrange for your traffic shaping configuration script to be run at that time. How you do that is distribution dependent and will not be covered here. You then should: Set TC_ENABLED=No and CLEAR_TC=No If your script uses the fwmark classifier, you can mark packets using entries in /etc/shorewall/tcrules.
Testing Tools At least one Shorewall user has found this tool helpful: http://e2epi.internet2.edu/network-performance-toolkit.html