$FreeBSD: /repoman/r/ncvs/src/sbin/ipfw/ipfw.8,v 1.63.2.38 2003/07/28 07:15:13 luigi Exp $
.Dd August 13, 2002 .Dt DNCTL 8 .Os Darwin .Sh NAME .Nm dnctl .Nd Traffic shaper control program .Sh SYNOPSIS .Nm .Op Fl anqs rq Cm list | show .Nm .Op Fl f | q .Cm flush .Nm .Op Fl q rq Cm delete .Op Ar number ... .Nm rq Cm pipe | queue .Ar number .Cm config .Ar config-options .Nm .Op Fl s Op Ar field rq Cm pipe | queue rq Cm delete | list | show .Op Ar number ... .Nm .Op Fl nq .Oo .Fl p Ar preproc .Oo .Ar preproc-flags .Oc .Oc .Ar pathname .Sh DESCRIPTION
p The .Nm utility is the user interface for controlling the .Xr dummynet 4 traffic shaper.
p .Nm dummynet operates by first using a packet filter to classify packets and divide them into .Em flows , using any match pattern that can be used in .Nm rules. Depending on local policies, a flow can contain packets for a single TCP connection, or from/to a given host, or entire subnet, or a protocol type, etc.
p Packets belonging to the same flow are then passed to either of two different objects, which implement the traffic regulation: l -hang -offset XXXX t Em pipe A pipe emulates a link with given bandwidth, propagation delay, queue size and packet loss rate. Packets are queued in front of the pipe as they come out from the classifier, and then transferred to the pipe according to the pipe's parameters.
p t Em queue A queue is an abstraction used to implement the WF2Q+ (Worst-case Fair Weighted Fair Queueing) policy, which is an efficient variant of the WFQ policy.
The queue associates a .Em weight and a reference pipe to each flow, and then all backlogged (i.e., with packets queued) flows linked to the same pipe share the pipe's bandwidth proportionally to their weights. Note that weights are not priorities; a flow with a lower weight is still guaranteed to get its fraction of the bandwidth even if a flow with a higher weight is permanently backlogged.
p .El In practice, .Em pipes can be used to set hard limits to the bandwidth that a flow can use, whereas .Em queues can be used to determine how different flow share the available bandwidth.
p The .Em pipe and .Em queue configuration commands are the following: d -ragged -offset indent .Cm pipe Ar number Cm config Ar pipe-configuration
p .Cm queue Ar number Cm config Ar queue-configuration .Ed
p The following parameters can be configured for a pipe:
p l -tag -width indent -compact t Cm bw Ar bandwidth | device Bandwidth, measured in .Sm off .Op Cm K | M rq Cm bit/s | Byte/s . .Sm on
p A value of 0 (default) means unlimited bandwidth. The unit must immediately follow the number, as in
p .Dl "dnctl pipe 1 config bw 300Kbit/s"
p If a device name is specified instead of a numeric value, as in
p .Dl "dnctl pipe 1 config bw tun0"
p then the transmit clock is supplied by the specified device. At the moment no device supports this functionality.
p t Cm delay Ar ms-delay Propagation delay, measured in milliseconds. The value is rounded to the next multiple of the clock tick (typically 10ms, but it is a good practice to run kernels with .Dq "options HZ=1000" to reduce the granularity to 1ms or less). Default value is 0, meaning no delay. .El
p The following parameters can be configured for a queue:
p l -tag -width indent -compact t Cm pipe Ar pipe_nr Connects a queue to the specified pipe. Multiple queues (with the same or different weights) can be connected to the same pipe, which specifies the aggregate rate for the set of queues.
p t Cm weight Ar weight Specifies the weight to be used for flows matching this queue. The weight must be in the range 1..100, and defaults to 1. .El
p Finally, the following parameters can be configured for both pipes and queues:
p l -tag -width XXXX -compact t Cm buckets Ar hash-table-size Specifies the size of the hash table used for storing the various queues. Default value is 64 controlled by the .Xr sysctl 8 variable .Em net.inet.ip.dummynet.hash_size , allowed range is 16 to 65536.
p t Cm mask Ar mask-specifier Packets sent to a given pipe or queue by an .Nm rule can be further classified into multiple flows, each of which is then sent to a different .Em dynamic pipe or queue. A flow identifier is constructed by masking the IP addresses, ports and protocol types as specified with the .Cm mask options in the configuration of the pipe or queue. For each different flow identifier, a new pipe or queue is created with the same parameters as the original object, and matching packets are sent to it.
p Thus, when .Em dynamic pipes are used, each flow will get the same bandwidth as defined by the pipe, whereas when .Em dynamic queues are used, each flow will share the parent's pipe bandwidth evenly with other flows generated by the same queue (note that other queues with different weights might be connected to the same pipe).
Available mask specifiers are a combination of one or more of the following:
p .Cm dst-ip Ar mask , .Cm dst-ip6 Ar mask , .Cm src-ip Ar mask , .Cm src-ip6 Ar mask , .Cm dst-port Ar mask , .Cm src-port Ar mask , .Cm proto Ar mask or .Cm all ,
p where the latter means all bits in all fields are significant.
p t Cm noerror When a packet is dropped by a dummynet queue or pipe, the error is normally reported to the caller routine in the kernel, in the same way as it happens when a device queue fills up. Setting this option reports the packet as successfully delivered, which can be needed for some experimental setups where you want to simulate loss or congestion at a remote router.
p t Cm plr Ar packet-loss-rate Packet loss rate. Argument .Ar packet-loss-rate is a floating-point number between 0 and 1, with 0 meaning no loss, 1 meaning 100% loss. The loss rate is internally represented on 31 bits.
p t Cm queue Brq Ar slots | size Ns Cm Kbytes Queue size, in .Ar slots or .Cm KBytes . Default value is 50 slots, which is the typical queue size for Ethernet devices. Note that for slow speed links you should keep the queue size short or your traffic might be affected by a significant queueing delay. E.g., 50 max-sized ethernet packets (1500 bytes) mean 600Kbit or 20s of queue on a 30Kbit/s pipe. Even worse effect can result if you get packets from an interface with a much larger MTU, e.g. the loopback interface with its 16KB packets.
p t Cm red | gred Ar w_q Ns / Ns Ar min_th Ns / Ns Ar max_th Ns / Ns Ar max_p Make use of the RED (Random Early Detection) queue management algorithm. .Ar w_q and .Ar max_p are floating point numbers between 0 and 1 (0 not included), while .Ar min_th and .Ar max_th are integer numbers specifying thresholds for queue management (thresholds are computed in bytes if the queue has been defined in bytes, in slots otherwise). The .Xr dummynet 4 also supports the gentle RED variant (gred).
p Three .Xr sysctl 8 variables can be used to control the RED behaviour: l -tag -width indent t Em net.inet.ip.dummynet.red_lookup_depth specifies the accuracy in computing the average queue when the link is idle (defaults to 256, must be greater than zero) t Em net.inet.ip.dummynet.red_avg_pkt_size specifies the expected average packet size (defaults to 512, must be greater than zero) t Em net.inet.ip.dummynet.red_max_pkt_size specifies the expected maximum packet size, only used when queue thresholds are in bytes (defaults to 1500, must be greater than zero). .El .El
p The following options are available: l -tag -width indent t Fl a While listing, show counter values. The .Cm show command just implies this option. t Fl f Don't ask for confirmation for commands that can cause problems if misused, .No i.e. Cm flush . If there is no tty associated with the process, this is implied. t Fl h Displays a short help. t Fl n Only check syntax of the command strings, without actually passing them to the kernel. t Fl q While .Cm add Ns ing , .Cm zero Ns ing , .Cm resetlog Ns ging or .Cm flush Ns ing , be quiet about actions (implies .Fl f ) . This is useful for adjusting rules by executing multiple .Nm commands in a script or by processing a file of many .Nm rules across a remote login session. If a .Cm flush is performed in normal (verbose) mode (with the default kernel configuration), it prints a message. Because all rules are flushed, the message might not be delivered to the login session, causing the remote login session to be closed and the remainder of the ruleset to not be processed. Access to the console would then be required to recover. t Fl s Op Ar field While listing pipes, sort according to one of the four counters (total or current packets or bytes). t Fl v Be verbose. .El
p To ease configuration, rules can be put into a file which is processed using .Nm as shown in the last synopsis line. An absolute .Ar pathname must be used. The file will be read line by line and applied as arguments to the .Nm utility.
p Optionally, a preprocessor can be specified using .Fl p Ar preproc where .Ar pathname is to be piped through. Useful preprocessors include .Xr cpp 1 and .Xr m4 1 . If .Ar preproc doesn't start with a slash
q Ql / as its first character, the usual .Ev PATH name search is performed. Care should be taken with this in environments where not all file systems are mounted (yet) by the time .Nm is being run (e.g. when they are mounted over NFS). Once .Fl p has been specified, any additional arguments as passed on to the preprocessor for interpretation. This allows for flexible configuration files (like conditionalizing them on the local hostname) and the use of macros to centralize frequently required arguments like IP addresses. .El .Sh CHECKLIST Here are some important points to consider when designing your rules: l -bullet t Remember that you filter both packets going .Cm in and .Cm out . Most connections need packets going in both directions. t Remember to test very carefully. It is a good idea to be near the console when doing this. If you cannot be near the console, use an auto-recovery script such as the one in
a /usr/share/examples/ipfw/change_rules.sh . t Don't forget the loopback interface. .El .Sh SYSCTL VARIABLES A set of .Xr sysctl 8 variables controls the behaviour of the .Nm dummynet module. These are shown below together with their default value (but always check with the .Xr sysctl 8 command what value is actually in use) and meaning: l -tag -width indent t Em net.inet.ip.dummynet.expire : No 1 Lazily delete dynamic pipes/queue once they have no pending traffic. You can disable this by setting the variable to 0, in which case the pipes/queues will only be deleted when the threshold is reached. t Em net.inet.ip.dummynet.hash_size : No 64 Default size of the hash table used for dynamic pipes/queues. This value is used when no .Cm buckets option is specified when configuring a pipe/queue. t Em net.inet.ip.dummynet.max_chain_len : No 16 Target value for the maximum number of pipes/queues in a hash bucket. The product .Cm max_chain_len*hash_size is used to determine the threshold over which empty pipes/queues will be expired even when .Cm net.inet.ip.dummynet.expire=0 . t Em net.inet.ip.dummynet.red_lookup_depth : No 256 t Em net.inet.ip.dummynet.red_avg_pkt_size : No 512 t Em net.inet.ip.dummynet.red_max_pkt_size : No 1500 Parameters used in the computations of the drop probability for the RED algorithm. .El .Sh EXAMPLES The following rules show some of the applications of .Xr dummynet 4 for simulations and the like. The examples use the obsolete .Nm ipfw command for illustration only as the use of .Nm ipfw is discouraged.
p This rule drops random incoming packets with a probability of 5%:
p .Dl "ipfw add pipe 10 ip from any to any" .Dl "dnctl pipe 10 config plr 0.05"
p We can use pipes to artificially limit bandwidth, e.g. on a machine acting as a router, if we want to limit traffic from local clients on 192.168.2.0/24 we do:
p .Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out" .Dl "dnctl pipe 1 config bw 300Kbit/s queue 50KBytes"
p note that we use the .Cm out modifier so that the rule is not used twice. Remember in fact that .Nm rules are checked both on incoming and outgoing packets.
p Should we want to simulate a bidirectional link with bandwidth limitations, the correct way is the following:
p .Dl "ipfw add pipe 1 ip from any to any out" .Dl "ipfw add pipe 2 ip from any to any in" .Dl "dnctl pipe 1 config bw 64Kbit/s queue 10Kbytes" .Dl "dnctl pipe 2 config bw 64Kbit/s queue 10Kbytes"
p The above can be very useful, e.g. if you want to see how your fancy Web page will look for a residential user who is connected only through a slow link. You should not use only one pipe for both directions, unless you want to simulate a half-duplex medium (e.g. AppleTalk, Ethernet, IRDA). It is not necessary that both pipes have the same configuration, so we can also simulate asymmetric links.
p Should we want to verify network performance with the RED queue management algorithm:
p .Dl "ipfw add pipe 1 ip from any to any" .Dl "dnctl pipe 1 config bw 500Kbit/s queue 100 red 0.002/30/80/0.1"
p Another typical application of the traffic shaper is to introduce some delay in the communication. This can significantly affect applications which do a lot of Remote Procedure Calls, and where the round-trip-time of the connection often becomes a limiting factor much more than bandwidth:
p .Dl "ipfw add pipe 1 ip from any to any out" .Dl "ipfw add pipe 2 ip from any to any in" .Dl "dnctl pipe 1 config delay 250ms bw 1Mbit/s" .Dl "dnctl pipe 2 config delay 250ms bw 1Mbit/s"
p Per-flow queueing can be useful for a variety of purposes. A very simple one is counting traffic:
p .Dl "ipfw add pipe 1 tcp from any to any" .Dl "ipfw add pipe 1 udp from any to any" .Dl "ipfw add pipe 1 ip from any to any" .Dl "dnctl pipe 1 config mask all"
p The above set of rules will create queues (and collect statistics) for all traffic. Because the pipes have no limitations, the only effect is collecting statistics. Note that we need 3 rules, not just the last one, because when .Nm tries to match IP packets it will not consider ports, so we would not see connections on separate ports as different ones.
p A more sophisticated example is limiting the outbound traffic on a net with per-host limits, rather than per-network limits:
p .Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out" .Dl "ipfw add pipe 2 ip from any to 192.168.2.0/24 in" .Dl "dnctl pipe 1 config mask src-ip 0x000000ff bw 200Kbit/s queue 20Kbytes" .Dl "dnctl pipe 2 config mask dst-ip 0x000000ff bw 200Kbit/s queue 20Kbytes" .Ss SETS OF RULES To add a set of rules atomically, e.g. set 18:
p .Dl "ipfw set disable 18" .Dl "ipfw add NN set 18 ... # repeat as needed" .Dl "ipfw set enable 18"
p To delete a set of rules atomically the command is simply:
p .Dl "ipfw delete set 18"
p To test a ruleset and disable it and regain control if something goes wrong:
p .Dl "ipfw set disable 18" .Dl "ipfw add NN set 18 ... # repeat as needed" .Dl "ipfw set enable 18; echo done; sleep 30 && ipfw set disable 18"
p Here if everything goes well, you press control-C before the "sleep" terminates, and your ruleset will be left active. Otherwise, e.g. if you cannot access your box, the ruleset will be disabled after the sleep terminates thus restoring the previous situation. .Sh SEE ALSO .Xr cpp 1 , .Xr m4 1 , .Xr dummynet 4 , .Xr ip 4 , .Xr ipfirewall 4 , .Xr protocols 5 , .Xr services 5 , .Xr sysctl 8 .Sh AUTHORS .An Ugen J. S. Antsilevich , .An Poul-Henning Kamp , .An Alex Nash , .An Archie Cobbs , .An Luigi Rizzo .
p .An -nosplit API based upon code written by .An Daniel Boulet for BSDI.
p Work on .Xr dummynet 4 traffic shaper supported by Akamba Corp. .Sh HISTORY The .Nm utility first appeared in .Fx 2.0 . .Xr dummynet 4 was introduced in .Fx 2.2.8 . Stateful extensions were introduced in .Fx 4.0 . .Nm ipfw2 was introduced in Summer 2002.