9 .Nd User interface for firewall, traffic shaper, packet scheduler,
12 .Ss FIREWALL CONFIGURATION
21 .Op Ar rule | first-last ...
29 .Brq Cm delete | zero | resetlog
33 .Cm set Oo Cm disable Ar number ... Oc Op Cm enable Ar number ...
37 .Ar number Cm to Ar number
39 .Cm set swap Ar number number
45 .Brq Cm firewall | altq | one_pass | debug | verbose | dyn_keepalive
48 .Brq Cm firewall | altq | one_pass | debug | verbose | dyn_keepalive
51 .Cm table Ar number Cm add Ar addr Ns Oo / Ns Ar masklen Oc Op Ar value
53 .Cm table Ar number Cm delete Ar addr Ns Op / Ns Ar masklen
62 .Ss DUMMYNET CONFIGURATION (TRAFFIC SHAPER AND PACKET SCHEDULER)
64 .Brq Cm pipe | queue | sched
70 .Brq Cm pipe | queue | sched
71 .Brq Cm delete | list | show
93 utility is the user interface for controlling the
97 traffic shaper/packet scheduler, and the
98 in-kernel NAT services.
100 A firewall configuration, or
104 numbered from 1 to 65535.
105 Packets are passed to the firewall
106 from a number of different places in the protocol stack
107 (depending on the source and destination of the packet,
108 it is possible for the firewall to be
109 invoked multiple times on the same packet).
110 The packet passed to the firewall is compared
111 against each of the rules in the
114 (multiple rules with the same number are permitted, in which case
115 they are processed in order of insertion).
116 When a match is found, the action corresponding to the
117 matching rule is performed.
119 Depending on the action and certain system settings, packets
120 can be reinjected into the firewall at some rule after the
121 matching one for further processing.
123 A ruleset always includes a
125 rule (numbered 65535) which cannot be modified or deleted,
126 and matches all packets.
127 The action associated with the
133 depending on how the kernel is configured.
135 If the ruleset includes one or more rules with the
140 the firewall will have a
142 behaviour, i.e., upon a match it will create
144 i.e., rules that match packets with the same 5-tuple
145 (protocol, source and destination addresses and ports)
146 as the packet which caused their creation.
147 Dynamic rules, which have a limited lifetime, are checked
148 at the first occurrence of a
153 rule, and are typically used to open the firewall on-demand to
154 legitimate traffic only.
156 .Sx STATEFUL FIREWALL
159 Sections below for more information on the stateful behaviour of
162 All rules (including dynamic ones) have a few associated counters:
163 a packet count, a byte count, a log count and a timestamp
164 indicating the time of the last match.
165 Counters can be displayed or reset with
169 Each rule belongs to one of 32 different
173 commands to atomically manipulate sets, such as enable,
174 disable, swap sets, move all rules in a set to another
175 one, delete all rules in a set.
176 These can be useful to
177 install temporary configurations, or to test them.
180 for more information on
183 Rules can be added with the
185 command; deleted individually or in groups with the
187 command, and globally (except those in set 31) with the
189 command; displayed, optionally with the content of the
195 Finally, counters can be reset with the
202 The following general options are available when invoking
204 .Bl -tag -width indent
206 Show counter values when listing rules.
209 command implies this option.
211 Only show the action and the comment, not the body of a rule.
215 When entering or showing rules, print them in compact form,
216 i.e., omitting the "ip from any to any" string
217 when this does not carry any additional information.
219 When listing, show dynamic rules in addition to static ones.
223 is specified, also show expired dynamic rules.
225 Do not ask for confirmation for commands that can cause problems
228 If there is no tty associated with the process, this is implied.
230 When listing a table (see the
232 section below for more information on lookup tables), format values
234 By default, values are shown as integers.
236 Only check syntax of the command strings, without actually passing
239 Try to resolve addresses and service names in output.
241 Be quiet when executing the
251 This is useful when updating rulesets by executing multiple
255 .Ql sh\ /etc/rc.firewall ) ,
256 or by processing a file with many
258 rules across a remote login session.
259 It also stops a table add or delete
260 from failing if the entry already exists or is not present.
262 The reason why this option may be important is that
263 for some of these actions,
265 may print a message; if the action results in blocking the
266 traffic to the remote client,
267 the remote login session will be closed
268 and the rest of the ruleset will not be processed.
269 Access to the console would then be required to recover.
271 When listing rules, show the
273 each rule belongs to.
274 If this flag is not specified, disabled rules will not be
277 When listing pipes, sort according to one of the four
278 counters (total or current packets or bytes).
280 When listing, show last match timestamp converted with ctime().
282 When listing, show last match timestamp as seconds from the epoch.
283 This form can be more convenient for postprocessing by scripts.
285 .Ss LIST OF RULES AND PREPROCESSING
286 To ease configuration, rules can be put into a file which is
289 as shown in the last synopsis line.
293 The file will be read line by line and applied as arguments to the
297 Optionally, a preprocessor can be specified using
301 is to be piped through.
302 Useful preprocessors include
308 does not start with a slash
310 as its first character, the usual
312 name search is performed.
313 Care should be taken with this in environments where not all
314 file systems are mounted (yet) by the time
316 is being run (e.g.\& when they are mounted over NFS).
319 has been specified, any additional arguments are passed on to the preprocessor
321 This allows for flexible configuration files (like conditionalizing
322 them on the local hostname) and the use of macros to centralize
323 frequently required arguments like IP addresses.
324 .Ss TRAFFIC SHAPER CONFIGURATION
330 commands are used to configure the traffic shaper and packet scheduler.
332 .Sx TRAFFIC SHAPER (DUMMYNET) CONFIGURATION
333 Section below for details.
335 If the world and the kernel get out of sync the
337 ABI may break, preventing you from being able to add any rules.
339 adversely effect the booting process.
344 to temporarily disable the firewall to regain access to the network,
345 allowing you to fix the problem.
347 A packet is checked against the active ruleset in multiple places
348 in the protocol stack, under control of several sysctl variables.
349 These places and variables are shown below, and it is important to
350 have this picture in mind in order to design a correct ruleset.
351 .Bd -literal -offset indent
354 +----------->-----------+
356 [ip(6)_input] [ip(6)_output] net.inet(6).ip(6).fw.enable=1
359 [ether_demux] [ether_output_frame] net.link.ether.ipfw=1
361 +-->--[bdg_forward]-->--+ net.link.bridge.ipfw=1
367 times the same packet goes through the firewall can
368 vary between 0 and 4 depending on packet source and
369 destination, and system configuration.
371 Note that as packets flow through the stack, headers can be
372 stripped or added to it, and so they may or may not be available
374 E.g., incoming packets will include the MAC header when
378 but the same packets will have the MAC header stripped off when
385 Also note that each packet is always checked against the complete ruleset,
386 irrespective of the place where the check occurs, or the source of the packet.
387 If a rule contains some match patterns or actions which are not valid
388 for the place of invocation (e.g.\& trying to match a MAC header within
392 the match pattern will not match, but a
394 operator in front of such patterns
398 match on those packets.
399 It is thus the responsibility of
400 the programmer, if necessary, to write a suitable ruleset to
401 differentiate among the possible places.
403 rules can be useful here, as an example:
404 .Bd -literal -offset indent
405 # packets from ether_demux or bdg_forward
406 ipfw add 10 skipto 1000 all from any to any layer2 in
407 # packets from ip_input
408 ipfw add 10 skipto 2000 all from any to any not layer2 in
409 # packets from ip_output
410 ipfw add 10 skipto 3000 all from any to any not layer2 out
411 # packets from ether_output_frame
412 ipfw add 10 skipto 4000 all from any to any layer2 out
415 (yes, at the moment there is no way to differentiate between
416 ether_demux and bdg_forward).
418 In general, each keyword or argument must be provided as
419 a separate command line argument, with no leading or trailing
421 Keywords are case-sensitive, whereas arguments may
422 or may not be case-sensitive depending on their nature
423 (e.g.\& uid's are, hostnames are not).
425 Some arguments (e.g., port or address lists) are comma-separated
427 In this case, spaces after commas ',' are allowed to make
428 the line more readable.
429 You can also put the entire
430 command (including flags) into a single argument.
431 E.g., the following forms are equivalent:
432 .Bd -literal -offset indent
433 ipfw -q add deny src-ip 10.0.0.0/24,127.0.0.1/8
434 ipfw -q add deny src-ip 10.0.0.0/24, 127.0.0.1/8
435 ipfw "-q add deny src-ip 10.0.0.0/24, 127.0.0.1/8"
438 The format of firewall rules is the following:
439 .Bd -ragged -offset indent
442 .Op Cm set Ar set_number
443 .Op Cm prob Ar match_probability
445 .Op Cm log Op Cm logamount Ar number
455 where the body of the rule specifies which information is used
456 for filtering packets, among the following:
458 .Bl -tag -width "Source and dest. addresses and ports" -offset XXX -compact
459 .It Layer-2 header fields
461 .It IPv4 and IPv6 Protocol
463 .It Source and dest. addresses and ports
467 .It Transmit and receive interface
469 .It Misc. IP header fields
470 Version, type of service, datagram length, identification,
471 fragment flag (non-zero IP offset),
474 .It IPv6 Extension headers
475 Fragmentation, Hop-by-Hop options,
476 Routing Headers, Source routing rthdr0, Mobile IPv6 rthdr2, IPSec options.
478 .It Misc. TCP header fields
479 TCP flags (SYN, FIN, ACK, RST, etc.),
480 sequence number, acknowledgment number,
488 When the packet can be associated with a local socket.
490 Whether a packet came from a divert socket (e.g.,
492 .It Fib annotation state
493 Whether a packet has been tagged for using a specific FIB (routing table)
494 in future forwarding decisions.
497 Note that some of the above information, e.g.\& source MAC or IP addresses and
498 TCP/UDP ports, can be easily spoofed, so filtering on those fields
499 alone might not guarantee the desired results.
500 .Bl -tag -width indent
502 Each rule is associated with a
504 in the range 1..65535, with the latter reserved for the
507 Rules are checked sequentially by rule number.
508 Multiple rules can have the same number, in which case they are
509 checked (and listed) according to the order in which they have
511 If a rule is entered without specifying a number, the kernel will
512 assign one in such a way that the rule becomes the last one
516 Automatic rule numbers are assigned by incrementing the last
517 non-default rule number by the value of the sysctl variable
518 .Ar net.inet.ip.fw.autoinc_step
519 which defaults to 100.
520 If this is not possible (e.g.\& because we would go beyond the
521 maximum allowed rule number), the number of the last
522 non-default value is used instead.
523 .It Cm set Ar set_number
524 Each rule is associated with a
527 Sets can be individually disabled and enabled, so this parameter
528 is of fundamental importance for atomic ruleset manipulation.
529 It can be also used to simplify deletion of groups of rules.
530 If a rule is entered without specifying a set number,
533 Set 31 is special in that it cannot be disabled,
534 and rules in set 31 are not deleted by the
536 command (but you can delete them with the
537 .Nm ipfw delete set 31
539 Set 31 is also used for the
542 .It Cm prob Ar match_probability
543 A match is only declared with the specified probability
544 (floating point number between 0 and 1).
545 This can be useful for a number of applications such as
546 random packet drop or
549 to simulate the effect of multiple paths leading to out-of-order
552 Note: this condition is checked before any other condition, including
553 ones such as keep-state or check-state which might have side effects.
554 .It Cm log Op Cm logamount Ar number
555 Packets matching a rule with the
557 keyword will be made available for logging in two ways:
558 if the sysctl variable
559 .Va net.inet.ip.fw.verbose
560 is set to 0 (default), one can use
565 This pseudo interface can be created after a boot
566 manually by using the following command:
567 .Bd -literal -offset indent
568 # ifconfig ipfw0 create
571 Or, automatically at boot time by adding the following
575 .Bd -literal -offset indent
579 There is no overhead if no
581 is attached to the pseudo interface.
584 .Va net.inet.ip.fw.verbose
585 is set to 1, packets will be logged to
589 facility up to a maximum of
594 is specified, the limit is taken from the sysctl variable
595 .Va net.inet.ip.fw.verbose_limit .
596 In both cases, a value of 0 means unlimited logging.
598 Once the limit is reached, logging can be re-enabled by
599 clearing the logging counter or the packet counter for that entry, see the
603 Note: logging is done after all other packet matching conditions
604 have been successfully verified, and before performing the final
605 action (accept, deny, etc.) on the packet.
607 When a packet matches a rule with the
609 keyword, the numeric tag for the given
611 in the range 1..65534 will be attached to the packet.
612 The tag acts as an internal marker (it is not sent out over
613 the wire) that can be used to identify these packets later on.
614 This can be used, for example, to provide trust between interfaces
615 and to start doing policy-based filtering.
616 A packet can have multiple tags at the same time.
617 Tags are "sticky", meaning once a tag is applied to a packet by a
618 matching rule it exists until explicit removal.
619 Tags are kept with the packet everywhere within the kernel, but are
620 lost when packet leaves the kernel, for example, on transmitting
621 packet out to the network or sending packet to a
625 To check for previously applied tags, use the
628 To delete previously applied tag, use the
632 Note: since tags are kept with the packet everywhere in kernelspace,
633 they can be set and unset anywhere in the kernel network subsystem
636 facility), not only by means of the
642 For example, there can be a specialized
644 node doing traffic analyzing and tagging for later inspecting
646 .It Cm untag Ar number
647 When a packet matches a rule with the
649 keyword, the tag with the number
651 is searched among the tags attached to this packet and,
652 if found, removed from it.
653 Other tags bound to packet, if present, are left untouched.
655 When a packet matches a rule with the
657 keyword, the ALTQ identifier for the given
662 Note that this ALTQ tag is only meaningful for packets going "out" of IPFW,
663 and not being rejected or going to divert sockets.
664 Note that if there is insufficient memory at the time the packet is
665 processed, it will not be tagged, so it is wise to make your ALTQ
666 "default" queue policy account for this.
669 rules match a single packet, only the first one adds the ALTQ classification
671 In doing so, traffic may be shaped by using
672 .Cm count Cm altq Ar queue
673 rules for classification early in the ruleset, then later applying
674 the filtering decision.
679 rules may come later and provide the actual filtering decisions in
680 addition to the fallback ALTQ tag.
684 to set up the queues before IPFW will be able to look them up by name,
685 and if the ALTQ disciplines are rearranged, the rules in containing the
686 queue identifiers in the kernel will likely have gone stale and need
688 Stale queue identifiers will probably result in misclassification.
690 All system ALTQ processing can be turned on or off via
695 .Cm disable Ar altq .
697 .Va net.inet.ip.fw.one_pass
698 is irrelevant to ALTQ traffic shaping, as the actual rule action is followed
699 always after adding an ALTQ tag.
702 A rule can be associated with one of the following actions, which
703 will be executed when the packet matches the body of the rule.
704 .Bl -tag -width indent
705 .It Cm allow | accept | pass | permit
706 Allow packets that match rule.
707 The search terminates.
709 Checks the packet against the dynamic ruleset.
710 If a match is found, execute the action associated with
711 the rule which generated this dynamic rule, otherwise
712 move to the next rule.
715 rules do not have a body.
718 rule is found, the dynamic ruleset is checked at the first
724 Update counters for all packets that match rule.
725 The search continues with the next rule.
727 Discard packets that match this rule.
728 The search terminates.
729 .It Cm divert Ar port
730 Divert packets that match this rule to the
734 The search terminates.
735 .It Cm fwd | forward Ar ipaddr | tablearg Ns Op , Ns Ar port
736 Change the next-hop on matching packets to
738 which can be an IP address or a host name.
739 For IPv4, the next hop can also be supplied by the last table
740 looked up for the packet by using the
742 keyword instead of an explicit address.
743 The search terminates if this rule matches.
747 is a local address, then matching packets will be forwarded to
749 (or the port number in the packet if one is not specified in the rule)
750 on the local machine.
754 is not a local address, then the port number
755 (if specified) is ignored, and the packet will be
756 forwarded to the remote address, using the route as found in
757 the local routing table for that IP.
761 rule will not match layer-2 packets (those received
762 on ether_input, ether_output, or bridged).
766 action does not change the contents of the packet at all.
767 In particular, the destination address remains unmodified, so
768 packets forwarded to another system will usually be rejected by that system
769 unless there is a matching rule on that system to capture them.
770 For packets forwarded locally,
771 the local address of the socket will be
772 set to the original destination address of the packet.
775 entry look rather weird but is intended for
776 use with transparent proxy servers.
777 .It Cm nat Ar nat_nr | tablearg
780 (for network address translation, address redirect, etc.):
782 .Sx NETWORK ADDRESS TRANSLATION (NAT)
783 Section for further information.
784 .It Cm pipe Ar pipe_nr
788 (for bandwidth limitation, delay, etc.).
790 .Sx TRAFFIC SHAPER (DUMMYNET) CONFIGURATION
791 Section for further information.
792 The search terminates; however, on exit from the pipe and if
796 .Va net.inet.ip.fw.one_pass
797 is not set, the packet is passed again to the firewall code
798 starting from the next rule.
799 .It Cm queue Ar queue_nr
803 (for bandwidth limitation using WF2Q+).
809 Discard packets that match this rule, and if the
810 packet is a TCP packet, try to send a TCP reset (RST) notice.
811 The search terminates.
813 Discard packets that match this rule, and if the
814 packet is a TCP packet, try to send a TCP reset (RST) notice.
815 The search terminates.
816 .It Cm skipto Ar number | tablearg
817 Skip all subsequent rules numbered less than
819 The search continues with the first rule numbered
822 It is possible to use the
824 keyword with a skipto for a
826 skipto, but care should be used, as no destination caching
827 is possible in this case so the rules are always walked to find it,
830 .It Cm call Ar number | tablearg
831 The current rule number is saved in the internal stack and
832 ruleset processing continues with the first rule numbered
835 If later a rule with the
837 action is encountered, the processing returns to the first rule
840 rule plus one or higher
841 (the same behaviour as with packets returning from
846 This could be used to make somewhat like an assembly language
848 calls to rules with common checks for different interfaces, etc.
850 Rule with any number could be called, not just forward jumps as with
852 So, to prevent endless loops in case of mistakes, both
856 actions don't do any jumps and simply go to the next rule if memory
857 can't be allocated or stack overflowed/undeflowed.
859 Internally stack for rule numbers is implemented using
861 facility and currently has size of 16 entries.
862 As mbuf tags are lost when packet leaves the kernel,
864 should not be used in subroutines to avoid endless loops
865 and other undesired effects.
867 Takes rule number saved to internal stack by the last
869 action and returns ruleset processing to the first rule
870 with number greater than number of corresponding
873 See description of the
875 action for more details.
881 and thus are unconditional, but
883 command-line utility currently requires every action except
886 While it is sometimes useful to return only on some packets,
887 usually you want to print just
890 A workaround for this is to use new syntax and
893 .Bd -literal -offset indent
894 # Add a rule without actual body
895 ipfw add 2999 return via any
897 # List rules without "from any to any" part
901 This cosmetic annoyance may be fixed in future releases.
903 Send a copy of packets matching this rule to the
907 The search continues with the next rule.
908 .It Cm unreach Ar code
909 Discard packets that match this rule, and try to send an ICMP
910 unreachable notice with code
914 is a number from 0 to 255, or one of these aliases:
915 .Cm net , host , protocol , port ,
916 .Cm needfrag , srcfail , net-unknown , host-unknown ,
917 .Cm isolated , net-prohib , host-prohib , tosnet ,
918 .Cm toshost , filter-prohib , host-precedence
920 .Cm precedence-cutoff .
921 The search terminates.
922 .It Cm unreach6 Ar code
923 Discard packets that match this rule, and try to send an ICMPv6
924 unreachable notice with code
928 is a number from 0, 1, 3 or 4, or one of these aliases:
929 .Cm no-route, admin-prohib, address
932 The search terminates.
933 .It Cm netgraph Ar cookie
934 Divert packet into netgraph with given
936 The search terminates.
937 If packet is later returned from netgraph it is either
938 accepted or continues with the next rule, depending on
939 .Va net.inet.ip.fw.one_pass
941 .It Cm ngtee Ar cookie
942 A copy of packet is diverted into netgraph, original
943 packet continues with the next rule.
946 for more information on
951 .It Cm setfib Ar fibnum | tablearg
952 The packet is tagged so as to use the FIB (routing table)
954 in any subsequent forwarding decisions.
955 In the current implementation, this is limited to the values 0 through 15, see
957 Processing continues at the next rule.
958 It is possible to use the
961 If the tablearg value is not within the compiled range of fibs,
962 the packet's fib is set to 0.
964 Queue and reassemble IP fragments.
965 If the packet is not fragmented, counters are updated and
966 processing continues with the next rule.
967 If the packet is the last logical fragment, the packet is reassembled and, if
968 .Va net.inet.ip.fw.one_pass
969 is set to 0, processing continues with the next rule.
970 Otherwise, the packet is allowed to pass and the search terminates.
971 If the packet is a fragment in the middle of a logical group of fragments,
973 processing stops immediately.
975 Fragment handling can be tuned via
976 .Va net.inet.ip.maxfragpackets
978 .Va net.inet.ip.maxfragsperpacket
979 which limit, respectively, the maximum number of processable
980 fragments (default: 800) and
981 the maximum number of fragments per packet (default: 16).
983 NOTA BENE: since fragments do not contain port numbers,
984 they should be avoided with the
987 Alternatively, direction-based (like
991 ) and source-based (like
993 ) match patterns can be used to select fragments.
995 Usually a simple rule like:
996 .Bd -literal -offset indent
997 # reassemble incoming fragments
998 ipfw add reass all from any to any in
1001 is all you need at the beginning of your ruleset.
1004 The body of a rule contains zero or more patterns (such as
1005 specific source and destination addresses or ports,
1006 protocol options, incoming or outgoing interfaces, etc.)
1007 that the packet must match in order to be recognised.
1008 In general, the patterns are connected by (implicit)
1010 operators -- i.e., all must match in order for the
1012 Individual patterns can be prefixed by the
1014 operator to reverse the result of the match, as in
1016 .Dl "ipfw add 100 allow ip from not 1.2.3.4 to any"
1018 Additionally, sets of alternative match patterns
1020 can be constructed by putting the patterns in
1021 lists enclosed between parentheses ( ) or braces { }, and
1024 operator as follows:
1026 .Dl "ipfw add 100 allow ip from { x or not y or z } to any"
1028 Only one level of parentheses is allowed.
1029 Beware that most shells have special meanings for parentheses
1030 or braces, so it is advisable to put a backslash \\ in front of them
1031 to prevent such interpretations.
1033 The body of a rule must in general include a source and destination
1037 can be used in various places to specify that the content of
1038 a required field is irrelevant.
1040 The rule body has the following format:
1041 .Bd -ragged -offset indent
1042 .Op Ar proto Cm from Ar src Cm to Ar dst
1046 The first part (proto from src to dst) is for backward
1047 compatibility with earlier versions of
1051 any match pattern (including MAC headers, IP protocols,
1052 addresses and ports) can be specified in the
1056 Rule fields have the following meaning:
1057 .Bl -tag -width indent
1058 .It Ar proto : protocol | Cm { Ar protocol Cm or ... }
1059 .It Ar protocol : Oo Cm not Oc Ar protocol-name | protocol-number
1060 An IP protocol specified by number or name
1061 (for a complete list see
1062 .Pa /etc/protocols ) ,
1063 or one of the following keywords:
1064 .Bl -tag -width indent
1066 Matches IPv4 packets.
1068 Matches IPv6 packets.
1077 option will be treated as inner protocol.
1085 .Cm { Ar protocol Cm or ... }
1088 is provided for convenience only but its use is deprecated.
1089 .It Ar src No and Ar dst : Bro Cm addr | Cm { Ar addr Cm or ... } Brc Op Oo Cm not Oc Ar ports
1090 An address (or a list, see below)
1091 optionally followed by
1097 with multiple addresses) is provided for convenience only and
1098 its use is discouraged.
1099 .It Ar addr : Oo Cm not Oc Bro
1100 .Cm any | me | me6 |
1101 .Cm table Ns Pq Ar number Ns Op , Ns Ar value
1102 .Ar | addr-list | addr-set
1104 .Bl -tag -width indent
1106 matches any IP address.
1108 matches any IP address configured on an interface in the system.
1110 matches any IPv6 address configured on an interface in the system.
1111 The address list is evaluated at the time the packet is
1113 .It Cm table Ns Pq Ar number Ns Op , Ns Ar value
1114 Matches any IPv4 address for which an entry exists in the lookup table
1116 If an optional 32-bit unsigned
1118 is also specified, an entry will match only if it has this value.
1121 section below for more information on lookup tables.
1123 .It Ar addr-list : ip-addr Ns Op Ns , Ns Ar addr-list
1125 A host or subnet address specified in one of the following ways:
1126 .Bl -tag -width indent
1127 .It Ar numeric-ip | hostname
1128 Matches a single IPv4 address, specified as dotted-quad or a hostname.
1129 Hostnames are resolved at the time the rule is added to the firewall list.
1130 .It Ar addr Ns / Ns Ar masklen
1131 Matches all addresses with base
1133 (specified as an IP address, a network number, or a hostname)
1137 As an example, 1.2.3.4/25 or 1.2.3.0/25 will match
1138 all IP numbers from 1.2.3.0 to 1.2.3.127 .
1139 .It Ar addr Ns : Ns Ar mask
1140 Matches all addresses with base
1142 (specified as an IP address, a network number, or a hostname)
1145 specified as a dotted quad.
1146 As an example, 1.2.3.4:255.0.255.0 or 1.0.3.0:255.0.255.0 will match
1148 This form is advised only for non-contiguous
1150 It is better to resort to the
1151 .Ar addr Ns / Ns Ar masklen
1152 format for contiguous masks, which is more compact and less
1155 .It Ar addr-set : addr Ns Oo Ns / Ns Ar masklen Oc Ns Cm { Ns Ar list Ns Cm }
1156 .It Ar list : Bro Ar num | num-num Brc Ns Op Ns , Ns Ar list
1157 Matches all addresses with base address
1159 (specified as an IP address, a network number, or a hostname)
1160 and whose last byte is in the list between braces { } .
1161 Note that there must be no spaces between braces and
1162 numbers (spaces after commas are allowed).
1163 Elements of the list can be specified as single entries
1167 field is used to limit the size of the set of addresses,
1168 and can have any value between 24 and 32.
1170 it will be assumed as 24.
1172 This format is particularly useful to handle sparse address sets
1173 within a single rule.
1174 Because the matching occurs using a
1175 bitmask, it takes constant time and dramatically reduces
1176 the complexity of rulesets.
1178 As an example, an address specified as 1.2.3.4/24{128,35-55,89}
1179 or 1.2.3.0/24{128,35-55,89}
1180 will match the following IP addresses:
1182 1.2.3.128, 1.2.3.35 to 1.2.3.55, 1.2.3.89 .
1183 .It Ar addr6-list : ip6-addr Ns Op Ns , Ns Ar addr6-list
1185 A host or subnet specified one of the following ways:
1186 .Bl -tag -width indent
1187 .It Ar numeric-ip | hostname
1188 Matches a single IPv6 address as allowed by
1191 Hostnames are resolved at the time the rule is added to the firewall
1193 .It Ar addr Ns / Ns Ar masklen
1194 Matches all IPv6 addresses with base
1196 (specified as allowed by
1204 No support for sets of IPv6 addresses is provided because IPv6 addresses
1205 are typically random past the initial prefix.
1206 .It Ar ports : Bro Ar port | port Ns \&- Ns Ar port Ns Brc Ns Op , Ns Ar ports
1207 For protocols which support port numbers (such as TCP and UDP), optional
1209 may be specified as one or more ports or port ranges, separated
1210 by commas but no spaces, and an optional
1215 notation specifies a range of ports (including boundaries).
1219 may be used instead of numeric port values.
1220 The length of the port list is limited to 30 ports or ranges,
1221 though one can specify larger ranges by using an
1225 section of the rule.
1229 can be used to escape the dash
1231 character in a service name (from a shell, the backslash must be
1232 typed twice to avoid the shell itself interpreting it as an escape
1235 .Dl "ipfw add count tcp from any ftp\e\e-data-ftp to any"
1237 Fragmented packets which have a non-zero offset (i.e., not the first
1238 fragment) will never match a rule which has one or more port
1242 option for details on matching fragmented packets.
1244 .Ss RULE OPTIONS (MATCH PATTERNS)
1245 Additional match patterns can be used within
1247 Zero or more of these so-called
1249 can be present in a rule, optionally prefixed by the
1251 operand, and possibly grouped into
1254 The following match patterns can be used (listed in alphabetical order):
1255 .Bl -tag -width indent
1256 .It Cm // this is a comment.
1257 Inserts the specified text as a comment in the rule.
1258 Everything following // is considered as a comment and stored in the rule.
1259 You can have comment-only rules, which are listed as having a
1261 action followed by the comment.
1266 Matches only packets generated by a divert socket.
1267 .It Cm diverted-loopback
1268 Matches only packets coming from a divert socket back into the IP stack
1270 .It Cm diverted-output
1271 Matches only packets going from a divert socket back outward to the IP
1272 stack output for delivery.
1273 .It Cm dst-ip Ar ip-address
1274 Matches IPv4 packets whose destination IP is one of the address(es)
1275 specified as argument.
1276 .It Bro Cm dst-ip6 | dst-ipv6 Brc Ar ip6-address
1277 Matches IPv6 packets whose destination IP is one of the address(es)
1278 specified as argument.
1279 .It Cm dst-port Ar ports
1280 Matches IP packets whose destination port is one of the port(s)
1281 specified as argument.
1283 Matches TCP packets that have the RST or ACK bits set.
1284 .It Cm ext6hdr Ar header
1285 Matches IPv6 packets containing the extended header given by
1287 Supported headers are:
1293 any type of Routing Header
1295 Source routing Routing Header Type 0
1297 Mobile IPv6 Routing Header Type 2
1301 IPSec authentication headers
1303 and IPsec encapsulated security payload headers
1305 .It Cm fib Ar fibnum
1306 Matches a packet that has been tagged to use
1307 the given FIB (routing table) number.
1308 .It Cm flow-id Ar labels
1309 Matches IPv6 packets containing any of the flow labels given in
1312 is a comma separated list of numeric flow labels.
1314 Matches packets that are fragments and not the first
1315 fragment of an IP datagram.
1316 Note that these packets will not have
1317 the next protocol header (e.g.\& TCP, UDP) so options that look into
1318 these headers cannot match.
1320 Matches all TCP or UDP packets sent by or received for a
1324 may be specified by name or number.
1325 .It Cm jail Ar prisonID
1326 Matches all TCP or UDP packets sent by or received for the
1327 jail whos prison ID is
1329 .It Cm icmptypes Ar types
1330 Matches ICMP packets whose ICMP type is in the list
1332 The list may be specified as any combination of
1333 individual types (numeric) separated by commas.
1334 .Em Ranges are not allowed .
1335 The supported ICMP types are:
1339 destination unreachable
1347 router advertisement
1351 time-to-live exceeded
1363 address mask request
1365 and address mask reply
1367 .It Cm icmp6types Ar types
1368 Matches ICMP6 packets whose ICMP6 type is in the list of
1370 The list may be specified as any combination of
1371 individual types (numeric) separated by commas.
1372 .Em Ranges are not allowed .
1374 Matches incoming or outgoing packets, respectively.
1378 are mutually exclusive (in fact,
1382 .It Cm ipid Ar id-list
1383 Matches IPv4 packets whose
1385 field has value included in
1387 which is either a single value or a list of values or ranges
1388 specified in the same way as
1390 .It Cm iplen Ar len-list
1391 Matches IP packets whose total length, including header and data, is
1394 which is either a single value or a list of values or ranges
1395 specified in the same way as
1397 .It Cm ipoptions Ar spec
1398 Matches packets whose IPv4 header contains the comma separated list of
1399 options specified in
1401 The supported IP options are:
1404 (strict source route),
1406 (loose source route),
1408 (record packet route) and
1411 The absence of a particular option may be denoted
1414 .It Cm ipprecedence Ar precedence
1415 Matches IPv4 packets whose precedence field is equal to
1418 Matches packets that have IPSEC history associated with them
1419 (i.e., the packet comes encapsulated in IPSEC, the kernel
1420 has IPSEC support and IPSEC_FILTERTUNNEL option, and can correctly
1423 Note that specifying
1425 is different from specifying
1427 as the latter will only look at the specific IP protocol field,
1428 irrespective of IPSEC kernel support and the validity of the IPSEC data.
1430 Further note that this flag is silently ignored in kernels without
1432 It does not affect rule processing when given and the
1433 rules are handled as if with no
1436 .It Cm iptos Ar spec
1437 Matches IPv4 packets whose
1439 field contains the comma separated list of
1440 service types specified in
1442 The supported IP types of service are:
1445 .Pq Dv IPTOS_LOWDELAY ,
1447 .Pq Dv IPTOS_THROUGHPUT ,
1449 .Pq Dv IPTOS_RELIABILITY ,
1451 .Pq Dv IPTOS_MINCOST ,
1453 .Pq Dv IPTOS_ECN_CE .
1454 The absence of a particular type may be denoted
1457 .It Cm ipttl Ar ttl-list
1458 Matches IPv4 packets whose time to live is included in
1460 which is either a single value or a list of values or ranges
1461 specified in the same way as
1463 .It Cm ipversion Ar ver
1464 Matches IP packets whose IP version field is
1467 Upon a match, the firewall will create a dynamic rule, whose
1468 default behaviour is to match bidirectional traffic between
1469 source and destination IP/port using the same protocol.
1470 The rule has a limited lifetime (controlled by a set of
1472 variables), and the lifetime is refreshed every time a matching
1475 Matches only layer2 packets, i.e., those passed to
1477 from ether_demux() and ether_output_frame().
1478 .It Cm limit Bro Cm src-addr | src-port | dst-addr | dst-port Brc Ar N
1479 The firewall will only allow
1481 connections with the same
1482 set of parameters as specified in the rule.
1484 of source and destination addresses and ports can be
1487 only IPv4 flows are supported.
1488 .It Cm lookup Bro Cm dst-ip | dst-port | src-ip | src-port | uid | jail Brc Ar N
1489 Search an entry in lookup table
1491 that matches the field specified as argument.
1492 If not found, the match fails.
1493 Otherwise, the match succeeds and
1495 is set to the value extracted from the table.
1497 This option can be useful to quickly dispatch traffic based on
1498 certain packet fields.
1501 section below for more information on lookup tables.
1502 .It Cm { MAC | mac } Ar dst-mac src-mac
1503 Match packets with a given
1507 addresses, specified as the
1509 keyword (matching any MAC address), or six groups of hex digits
1510 separated by colons,
1511 and optionally followed by a mask indicating the significant bits.
1512 The mask may be specified using either of the following methods:
1513 .Bl -enum -width indent
1517 followed by the number of significant bits.
1518 For example, an address with 33 significant bits could be specified as:
1520 .Dl "MAC 10:20:30:40:50:60/33 any"
1525 followed by a bitmask specified as six groups of hex digits separated
1527 For example, an address in which the last 16 bits are significant could
1530 .Dl "MAC 10:20:30:40:50:60&00:00:00:00:ff:ff any"
1532 Note that the ampersand character has a special meaning in many shells
1533 and should generally be escaped.
1536 Note that the order of MAC addresses (destination first,
1538 the same as on the wire, but the opposite of the one used for
1540 .It Cm mac-type Ar mac-type
1541 Matches packets whose Ethernet Type field
1542 corresponds to one of those specified as argument.
1544 is specified in the same way as
1546 (i.e., one or more comma-separated single values or ranges).
1547 You can use symbolic names for known values such as
1548 .Em vlan , ipv4, ipv6 .
1549 Values can be entered as decimal or hexadecimal (if prefixed by 0x),
1550 and they are always printed as hexadecimal (unless the
1552 option is used, in which case symbolic resolution will be attempted).
1553 .It Cm proto Ar protocol
1554 Matches packets with the corresponding IP protocol.
1555 .It Cm recv | xmit | via Brq Ar ifX | Ar if Ns Cm * | Ar table Ns Pq Ar number Ns Op , Ns Ar value | Ar ipno | Ar any
1556 Matches packets received, transmitted or going through,
1557 respectively, the interface specified by exact name
1561 by IP address, or through some interface.
1565 keyword causes the interface to always be checked.
1572 then only the receive or transmit interface (respectively)
1574 By specifying both, it is possible to match packets based on
1575 both receive and transmit interface, e.g.:
1577 .Dl "ipfw add deny ip from any to any out recv ed0 xmit ed1"
1581 interface can be tested on either incoming or outgoing packets,
1584 interface can only be tested on outgoing packets.
1589 is invalid) whenever
1593 A packet might not have a receive or transmit interface: packets
1594 originating from the local host have no receive interface,
1595 while packets destined for the local host have no transmit
1598 Matches TCP packets that have the SYN bit set but no ACK bit.
1599 This is the short form of
1600 .Dq Li tcpflags\ syn,!ack .
1602 Matches packets that are associated to a local socket and
1603 for which the SO_USER_COOKIE socket option has been set
1604 to a non-zero value.
1605 As a side effect, the value of the
1606 option is made available as
1608 value, which in turn can be used as
1613 .It Cm src-ip Ar ip-address
1614 Matches IPv4 packets whose source IP is one of the address(es)
1615 specified as an argument.
1616 .It Cm src-ip6 Ar ip6-address
1617 Matches IPv6 packets whose source IP is one of the address(es)
1618 specified as an argument.
1619 .It Cm src-port Ar ports
1620 Matches IP packets whose source port is one of the port(s)
1621 specified as argument.
1622 .It Cm tagged Ar tag-list
1623 Matches packets whose tags are included in
1625 which is either a single value or a list of values or ranges
1626 specified in the same way as
1628 Tags can be applied to the packet using
1630 rule action parameter (see it's description for details on tags).
1631 .It Cm tcpack Ar ack
1633 Match if the TCP header acknowledgment number field is set to
1635 .It Cm tcpdatalen Ar tcpdatalen-list
1636 Matches TCP packets whose length of TCP data is
1637 .Ar tcpdatalen-list ,
1638 which is either a single value or a list of values or ranges
1639 specified in the same way as
1641 .It Cm tcpflags Ar spec
1643 Match if the TCP header contains the comma separated list of
1646 The supported TCP flags are:
1655 The absence of a particular flag may be denoted
1658 A rule which contains a
1660 specification can never match a fragmented packet which has
1664 option for details on matching fragmented packets.
1665 .It Cm tcpseq Ar seq
1667 Match if the TCP header sequence number field is set to
1669 .It Cm tcpwin Ar tcpwin-list
1670 Matches TCP packets whose header window field is set to
1672 which is either a single value or a list of values or ranges
1673 specified in the same way as
1675 .It Cm tcpoptions Ar spec
1677 Match if the TCP header contains the comma separated list of
1678 options specified in
1680 The supported TCP options are:
1683 (maximum segment size),
1685 (tcp window advertisement),
1689 (rfc1323 timestamp) and
1691 (rfc1644 t/tcp connection count).
1692 The absence of a particular option may be denoted
1696 Match all TCP or UDP packets sent by or received for a
1700 may be matched by name or identification number.
1702 For incoming packets,
1703 a routing table lookup is done on the packet's source address.
1704 If the interface on which the packet entered the system matches the
1705 outgoing interface for the route,
1707 If the interfaces do not match up,
1708 the packet does not match.
1709 All outgoing packets or packets with no incoming interface match.
1711 The name and functionality of the option is intentionally similar to
1712 the Cisco IOS command:
1714 .Dl ip verify unicast reverse-path
1716 This option can be used to make anti-spoofing rules to reject all
1717 packets with source addresses not from this interface.
1721 For incoming packets,
1722 a routing table lookup is done on the packet's source address.
1723 If a route to the source address exists, but not the default route
1724 or a blackhole/reject route, the packet matches.
1725 Otherwise, the packet does not match.
1726 All outgoing packets match.
1728 The name and functionality of the option is intentionally similar to
1729 the Cisco IOS command:
1731 .Dl ip verify unicast source reachable-via any
1733 This option can be used to make anti-spoofing rules to reject all
1734 packets whose source address is unreachable.
1736 For incoming packets, the packet's source address is checked if it
1737 belongs to a directly connected network.
1738 If the network is directly connected, then the interface the packet
1739 came on in is compared to the interface the network is connected to.
1740 When incoming interface and directly connected interface are not the
1741 same, the packet does not match.
1742 Otherwise, the packet does match.
1743 All outgoing packets match.
1745 This option can be used to make anti-spoofing rules to reject all
1746 packets that pretend to be from a directly connected network but do
1747 not come in through that interface.
1748 This option is similar to but more restricted than
1750 because it engages only on packets with source addresses of directly
1751 connected networks instead of all source addresses.
1754 Lookup tables are useful to handle large sparse sets of
1755 addresses or other search keys (e.g., ports, jail IDs, interface names).
1756 In the rest of this section we will use the term ``address''.
1757 There may be up to 65535 different lookup tables, numbered 0 to 65534.
1759 Each entry is represented by an
1760 .Ar addr Ns Op / Ns Ar masklen
1761 and will match all addresses with base
1763 (specified as an IPv4/IPv6 address, a hostname or an unsigned integer)
1769 is not specified, it defaults to 32 for IPv4 and 128 for IPv6.
1770 When looking up an IP address in a table, the most specific
1772 Associated with each entry is a 32-bit unsigned
1774 which can optionally be checked by a rule matching code.
1775 When adding an entry, if
1777 is not specified, it defaults to 0.
1779 An entry can be added to a table
1781 or removed from a table
1783 A table can be examined
1788 Internally, each table is stored in a Radix tree, the same way as
1789 the routing table (see
1792 Lookup tables currently support only ports, jail IDs, IPv4/IPv6 addresses
1793 and interface names.
1794 Wildcards is not supported for interface names.
1798 feature provides the ability to use a value, looked up in the table, as
1799 the argument for a rule action, action parameter or rule option.
1800 This can significantly reduce number of rules in some configurations.
1801 If two tables are used in a rule, the result of the second (destination)
1805 argument can be used with the following actions:
1806 .Cm nat, pipe , queue, divert, tee, netgraph, ngtee, fwd, skipto, setfib,
1814 it is possible to supply table entries with values
1815 that are in the form of IP addresses or hostnames.
1818 Section for example usage of tables and the tablearg keyword.
1822 action, the user should be aware that the code will walk the ruleset
1823 up to a rule equal to, or past, the given number,
1824 and should therefore try keep the
1825 ruleset compact between the skipto and the target rules.
1827 Each rule belongs to one of 32 different
1830 Set 31 is reserved for the default rule.
1832 By default, rules are put in set 0, unless you use the
1834 attribute when entering a new rule.
1835 Sets can be individually and atomically enabled or disabled,
1836 so this mechanism permits an easy way to store multiple configurations
1837 of the firewall and quickly (and atomically) switch between them.
1838 The command to enable/disable sets is
1839 .Bd -ragged -offset indent
1841 .Cm set Oo Cm disable Ar number ... Oc Op Cm enable Ar number ...
1848 sections can be specified.
1849 Command execution is atomic on all the sets specified in the command.
1850 By default, all sets are enabled.
1852 When you disable a set, its rules behave as if they do not exist
1853 in the firewall configuration, with only one exception:
1854 .Bd -ragged -offset indent
1855 dynamic rules created from a rule before it had been disabled
1856 will still be active until they expire.
1858 dynamic rules you have to explicitly delete the parent rule
1859 which generated them.
1862 The set number of rules can be changed with the command
1863 .Bd -ragged -offset indent
1866 .Brq Cm rule Ar rule-number | old-set
1870 Also, you can atomically swap two rulesets with the command
1871 .Bd -ragged -offset indent
1873 .Cm set swap Ar first-set second-set
1878 Section on some possible uses of sets of rules.
1879 .Sh STATEFUL FIREWALL
1880 Stateful operation is a way for the firewall to dynamically
1881 create rules for specific flows when packets that
1882 match a given pattern are detected.
1883 Support for stateful
1884 operation comes through the
1885 .Cm check-state , keep-state
1891 Dynamic rules are created when a packet matches a
1895 rule, causing the creation of a
1897 rule which will match all and only packets with
1901 .Em src-ip/src-port dst-ip/dst-port
1906 are used here only to denote the initial match addresses, but they
1907 are completely equivalent afterwards).
1908 Dynamic rules will be checked at the first
1909 .Cm check-state, keep-state
1912 occurrence, and the action performed upon a match will be the same
1913 as in the parent rule.
1915 Note that no additional attributes other than protocol and IP addresses
1916 and ports are checked on dynamic rules.
1918 The typical use of dynamic rules is to keep a closed firewall configuration,
1919 but let the first TCP SYN packet from the inside network install a
1920 dynamic rule for the flow so that packets belonging to that session
1921 will be allowed through the firewall:
1923 .Dl "ipfw add check-state"
1924 .Dl "ipfw add allow tcp from my-subnet to any setup keep-state"
1925 .Dl "ipfw add deny tcp from any to any"
1927 A similar approach can be used for UDP, where an UDP packet coming
1928 from the inside will install a dynamic rule to let the response through
1931 .Dl "ipfw add check-state"
1932 .Dl "ipfw add allow udp from my-subnet to any keep-state"
1933 .Dl "ipfw add deny udp from any to any"
1935 Dynamic rules expire after some time, which depends on the status
1936 of the flow and the setting of some
1940 .Sx SYSCTL VARIABLES
1942 For TCP sessions, dynamic rules can be instructed to periodically
1943 send keepalive packets to refresh the state of the rule when it is
1948 for more examples on how to use dynamic rules.
1949 .Sh TRAFFIC SHAPER (DUMMYNET) CONFIGURATION
1951 is also the user interface for the
1953 traffic shaper, packet scheduler and network emulator, a subsystem that
1954 can artificially queue, delay or drop packets
1955 emulating the behaviour of certain network links
1956 or queueing systems.
1959 operates by first using the firewall to select packets
1960 using any match pattern that can be used in
1963 Matching packets are then passed to either of two
1964 different objects, which implement the traffic regulation:
1965 .Bl -hang -offset XXXX
1971 with given bandwidth and propagation delay,
1972 driven by a FIFO scheduler and a single queue with programmable
1973 queue size and packet loss rate.
1974 Packets are appended to the queue as they come out from
1976 and then transferred in FIFO order to the link at the desired rate.
1980 is an abstraction used to implement packet scheduling
1981 using one of several packet scheduling algorithms.
1984 are first grouped into flows according to a mask on the 5-tuple.
1985 Flows are then passed to the scheduler associated to the
1987 and each flow uses scheduling parameters (weight and others)
1988 as configured in the
1991 A scheduler in turn is connected to an emulated link,
1992 and arbitrates the link's bandwidth among backlogged flows according to
1993 weights and to the features of the scheduling algorithm in use.
1998 can be used to set hard limits to the bandwidth that a flow can use, whereas
2000 can be used to determine how different flows share the available bandwidth.
2002 A graphical representation of the binding of queues,
2003 flows, schedulers and links is below.
2004 .Bd -literal -offset indent
2005 (flow_mask|sched_mask) sched_mask
2006 +---------+ weight Wx +-------------+
2007 | |->-[flow]-->--| |-+
2008 -->--| QUEUE x | ... | | |
2009 | |->-[flow]-->--| SCHEDuler N | |
2011 ... | +--[LINK N]-->--
2012 +---------+ weight Wy | | +--[LINK N]-->--
2013 | |->-[flow]-->--| | |
2014 -->--| QUEUE y | ... | | |
2015 | |->-[flow]-->--| | |
2016 +---------+ +-------------+ |
2019 It is important to understand the role of the SCHED_MASK
2020 and FLOW_MASK, which are configured through the commands
2021 .Dl "ipfw sched N config mask SCHED_MASK ..."
2023 .Dl "ipfw queue X config mask FLOW_MASK ..." .
2025 The SCHED_MASK is used to assign flows to one or more
2026 scheduler instances, one for each
2027 value of the packet's 5-tuple after applying SCHED_MASK.
2028 As an example, using ``src-ip 0xffffff00'' creates one instance
2029 for each /24 destination subnet.
2031 The FLOW_MASK, together with the SCHED_MASK, is used to split
2033 As an example, using
2034 ``src-ip 0x000000ff''
2035 together with the previous SCHED_MASK makes a flow for
2036 each individual source address.
2037 In turn, flows for each /24
2038 subnet will be sent to the same scheduler instance.
2040 The above diagram holds even for the
2042 case, with the only restriction that a
2044 only supports a SCHED_MASK, and forces the use of a FIFO
2045 scheduler (these are for backward compatibility reasons;
2046 in fact, internally, a
2048 pipe is implemented exactly as above).
2050 There are two modes of
2058 mode tries to emulate a real link: the
2060 scheduler ensures that the packet will not leave the pipe faster than it
2061 would on the real link with a given bandwidth.
2064 mode allows certain packets to bypass the
2066 scheduler (if packet flow does not exceed pipe's bandwidth).
2067 This is the reason why the
2069 mode requires less CPU cycles per packet (on average) and packet latency
2070 can be significantly lower in comparison to a real link with the same
2076 mode can be enabled by setting the
2077 .Va net.inet.ip.dummynet.io_fast
2079 variable to a non-zero value.
2081 .Ss PIPE, QUEUE AND SCHEDULER CONFIGURATION
2087 configuration commands are the following:
2088 .Bd -ragged -offset indent
2089 .Cm pipe Ar number Cm config Ar pipe-configuration
2091 .Cm queue Ar number Cm config Ar queue-configuration
2093 .Cm sched Ar number Cm config Ar sched-configuration
2096 The following parameters can be configured for a pipe:
2098 .Bl -tag -width indent -compact
2099 .It Cm bw Ar bandwidth | device
2100 Bandwidth, measured in
2103 .Brq Cm bit/s | Byte/s .
2106 A value of 0 (default) means unlimited bandwidth.
2107 The unit must immediately follow the number, as in
2109 .Dl "ipfw pipe 1 config bw 300Kbit/s"
2111 If a device name is specified instead of a numeric value, as in
2113 .Dl "ipfw pipe 1 config bw tun0"
2115 then the transmit clock is supplied by the specified device.
2116 At the moment only the
2118 device supports this
2119 functionality, for use in conjunction with
2122 .It Cm delay Ar ms-delay
2123 Propagation delay, measured in milliseconds.
2124 The value is rounded to the next multiple of the clock tick
2125 (typically 10ms, but it is a good practice to run kernels
2127 .Dq "options HZ=1000"
2129 the granularity to 1ms or less).
2130 The default value is 0, meaning no delay.
2132 .It Cm burst Ar size
2133 If the data to be sent exceeds the pipe's bandwidth limit
2134 (and the pipe was previously idle), up to
2136 bytes of data are allowed to bypass the
2138 scheduler, and will be sent as fast as the physical link allows.
2139 Any additional data will be transmitted at the rate specified
2143 The burst size depends on how long the pipe has been idle;
2144 the effective burst size is calculated as follows:
2151 .It Cm profile Ar filename
2152 A file specifying the additional overhead incurred in the transmission
2153 of a packet on the link.
2155 Some link types introduce extra delays in the transmission
2156 of a packet, e.g., because of MAC level framing, contention on
2157 the use of the channel, MAC level retransmissions and so on.
2158 From our point of view, the channel is effectively unavailable
2159 for this extra time, which is constant or variable depending
2161 Additionally, packets may be dropped after this
2162 time (e.g., on a wireless link after too many retransmissions).
2163 We can model the additional delay with an empirical curve
2164 that represents its distribution.
2165 .Bd -literal -offset indent
2166 cumulative probability
2176 +-------*------------------->
2179 The empirical curve may have both vertical and horizontal lines.
2180 Vertical lines represent constant delay for a range of
2182 Horizontal lines correspond to a discontinuity in the delay
2183 distribution: the pipe will use the largest delay for a
2186 The file format is the following, with whitespace acting as
2187 a separator and '#' indicating the beginning a comment:
2188 .Bl -tag -width indent
2189 .It Cm name Ar identifier
2190 optional name (listed by "ipfw pipe show")
2191 to identify the delay distribution;
2193 the bandwidth used for the pipe.
2194 If not specified here, it must be present
2195 explicitly as a configuration parameter for the pipe;
2196 .It Cm loss-level Ar L
2197 the probability above which packets are lost.
2198 (0.0 <= L <= 1.0, default 1.0 i.e., no loss);
2200 the number of samples used in the internal
2201 representation of the curve (2..1024; default 100);
2202 .It Cm "delay prob" | "prob delay"
2203 One of these two lines is mandatory and defines
2204 the format of the following lines with data points.
2206 2 or more lines representing points in the curve,
2207 with either delay or probability first, according
2208 to the chosen format.
2209 The unit for delay is milliseconds.
2210 Data points do not need to be sorted.
2211 Also, the number of actual lines can be different
2212 from the value of the "samples" parameter:
2214 utility will sort and interpolate
2215 the curve as needed.
2218 Example of a profile file:
2219 .Bd -literal -offset indent
2224 0 200 # minimum overhead is 200ms
2230 #configuration file end
2234 The following parameters can be configured for a queue:
2236 .Bl -tag -width indent -compact
2237 .It Cm pipe Ar pipe_nr
2238 Connects a queue to the specified pipe.
2239 Multiple queues (with the same or different weights) can be connected to
2240 the same pipe, which specifies the aggregate rate for the set of queues.
2242 .It Cm weight Ar weight
2243 Specifies the weight to be used for flows matching this queue.
2244 The weight must be in the range 1..100, and defaults to 1.
2247 The following case-insensitive parameters can be configured for a
2250 .Bl -tag -width indent -compact
2251 .It Cm type Ar {fifo | wf2q+ | rr | qfq}
2252 specifies the scheduling algorithm to use.
2253 .Bl -tag -width indent -compact
2255 is just a FIFO scheduler (which means that all packets
2256 are stored in the same queue as they arrive to the scheduler).
2257 FIFO has O(1) per-packet time complexity, with very low
2258 constants (estimate 60-80ns on a 2GHz desktop machine)
2259 but gives no service guarantees.
2261 implements the WF2Q+ algorithm, which is a Weighted Fair Queueing
2262 algorithm which permits flows to share bandwidth according to
2264 Note that weights are not priorities; even a flow
2265 with a minuscule weight will never starve.
2266 WF2Q+ has O(log N) per-packet processing cost, where N is the number
2267 of flows, and is the default algorithm used by previous versions
2270 implements the Deficit Round Robin algorithm, which has O(1) processing
2271 costs (roughly, 100-150ns per packet)
2272 and permits bandwidth allocation according to weights, but
2273 with poor service guarantees.
2275 implements the QFQ algorithm, which is a very fast variant of
2276 WF2Q+, with similar service guarantees and O(1) processing
2277 costs (roughly, 200-250ns per packet).
2281 In addition to the type, all parameters allowed for a pipe can also
2282 be specified for a scheduler.
2284 Finally, the following parameters can be configured for both
2287 .Bl -tag -width XXXX -compact
2288 .It Cm buckets Ar hash-table-size
2289 Specifies the size of the hash table used for storing the
2291 Default value is 64 controlled by the
2294 .Va net.inet.ip.dummynet.hash_size ,
2295 allowed range is 16 to 65536.
2297 .It Cm mask Ar mask-specifier
2298 Packets sent to a given pipe or queue by an
2300 rule can be further classified into multiple flows, each of which is then
2304 A flow identifier is constructed by masking the IP addresses,
2305 ports and protocol types as specified with the
2307 options in the configuration of the pipe or queue.
2308 For each different flow identifier, a new pipe or queue is created
2309 with the same parameters as the original object, and matching packets
2314 are used, each flow will get the same bandwidth as defined by the pipe,
2317 are used, each flow will share the parent's pipe bandwidth evenly
2318 with other flows generated by the same queue (note that other queues
2319 with different weights might be connected to the same pipe).
2321 Available mask specifiers are a combination of one or more of the following:
2323 .Cm dst-ip Ar mask ,
2324 .Cm dst-ip6 Ar mask ,
2325 .Cm src-ip Ar mask ,
2326 .Cm src-ip6 Ar mask ,
2327 .Cm dst-port Ar mask ,
2328 .Cm src-port Ar mask ,
2329 .Cm flow-id Ar mask ,
2334 where the latter means all bits in all fields are significant.
2337 When a packet is dropped by a
2339 queue or pipe, the error
2340 is normally reported to the caller routine in the kernel, in the
2341 same way as it happens when a device queue fills up.
2343 option reports the packet as successfully delivered, which can be
2344 needed for some experimental setups where you want to simulate
2345 loss or congestion at a remote router.
2347 .It Cm plr Ar packet-loss-rate
2350 .Ar packet-loss-rate
2351 is a floating-point number between 0 and 1, with 0 meaning no
2352 loss, 1 meaning 100% loss.
2353 The loss rate is internally represented on 31 bits.
2355 .It Cm queue Brq Ar slots | size Ns Cm Kbytes
2360 Default value is 50 slots, which
2361 is the typical queue size for Ethernet devices.
2362 Note that for slow speed links you should keep the queue
2363 size short or your traffic might be affected by a significant
2365 E.g., 50 max-sized ethernet packets (1500 bytes) mean 600Kbit
2366 or 20s of queue on a 30Kbit/s pipe.
2367 Even worse effects can result if you get packets from an
2368 interface with a much larger MTU, e.g.\& the loopback interface
2369 with its 16KB packets.
2373 .Em net.inet.ip.dummynet.pipe_byte_limit
2375 .Em net.inet.ip.dummynet.pipe_slot_limit
2376 control the maximum lengths that can be specified.
2378 .It Cm red | gred Ar w_q Ns / Ns Ar min_th Ns / Ns Ar max_th Ns / Ns Ar max_p
2379 Make use of the RED (Random Early Detection) queue management algorithm.
2384 point numbers between 0 and 1 (0 not included), while
2388 are integer numbers specifying thresholds for queue management
2389 (thresholds are computed in bytes if the queue has been defined
2390 in bytes, in slots otherwise).
2393 also supports the gentle RED variant (gred).
2396 variables can be used to control the RED behaviour:
2397 .Bl -tag -width indent
2398 .It Va net.inet.ip.dummynet.red_lookup_depth
2399 specifies the accuracy in computing the average queue
2400 when the link is idle (defaults to 256, must be greater than zero)
2401 .It Va net.inet.ip.dummynet.red_avg_pkt_size
2402 specifies the expected average packet size (defaults to 512, must be
2404 .It Va net.inet.ip.dummynet.red_max_pkt_size
2405 specifies the expected maximum packet size, only used when queue
2406 thresholds are in bytes (defaults to 1500, must be greater than zero).
2410 When used with IPv6 data,
2412 currently has several limitations.
2413 Information necessary to route link-local packets to an
2414 interface is not available after processing by
2416 so those packets are dropped in the output path.
2417 Care should be taken to ensure that link-local packets are not passed to
2420 Here are some important points to consider when designing your
2424 Remember that you filter both packets going
2428 Most connections need packets going in both directions.
2430 Remember to test very carefully.
2431 It is a good idea to be near the console when doing this.
2432 If you cannot be near the console,
2433 use an auto-recovery script such as the one in
2434 .Pa /usr/share/examples/ipfw/change_rules.sh .
2436 Do not forget the loopback interface.
2441 There are circumstances where fragmented datagrams are unconditionally
2443 TCP packets are dropped if they do not contain at least 20 bytes of
2444 TCP header, UDP packets are dropped if they do not contain a full 8
2445 byte UDP header, and ICMP packets are dropped if they do not contain
2446 4 bytes of ICMP header, enough to specify the ICMP type, code, and
2448 These packets are simply logged as
2450 since there may not be enough good data in the packet to produce a
2451 meaningful log entry.
2453 Another type of packet is unconditionally dropped, a TCP packet with a
2454 fragment offset of one.
2455 This is a valid packet, but it only has one use, to try
2456 to circumvent firewalls.
2457 When logging is enabled, these packets are
2458 reported as being dropped by rule -1.
2460 If you are logged in over a network, loading the
2464 is probably not as straightforward as you would think.
2465 The following command line is recommended:
2466 .Bd -literal -offset indent
2468 ipfw add 32000 allow ip from any to any
2471 Along the same lines, doing an
2472 .Bd -literal -offset indent
2476 in similar surroundings is also a bad idea.
2480 filter list may not be modified if the system security level
2481 is set to 3 or higher
2484 for information on system security levels).
2486 .Sh PACKET DIVERSION
2489 socket bound to the specified port will receive all packets
2490 diverted to that port.
2491 If no socket is bound to the destination port, or if the divert module is
2492 not loaded, or if the kernel was not compiled with divert socket support,
2493 the packets are dropped.
2494 .Sh NETWORK ADDRESS TRANSLATION (NAT)
2496 support in-kernel NAT using the kernel version of
2499 The nat configuration command is the following:
2500 .Bd -ragged -offset indent
2505 .Ar nat-configuration
2509 The following parameters can be configured:
2510 .Bl -tag -width indent
2511 .It Cm ip Ar ip_address
2512 Define an ip address to use for aliasing.
2514 Use ip address of NIC for aliasing, dynamically changing
2515 it if NIC's ip address changes.
2517 Enable logging on this nat instance.
2519 Deny any incoming connection from outside world.
2521 Try to leave the alias port numbers unchanged from
2522 the actual local port numbers.
2524 Traffic on the local network not originating from an
2525 unregistered address spaces will be ignored.
2527 Reset table of the packet aliasing engine on address change.
2529 Reverse the way libalias handles aliasing.
2531 Obey transparent proxy rules only, packet aliasing is not performed.
2533 Skip instance in case of global state lookup (see below).
2536 Some specials value can be supplied instead of
2538 .Bl -tag -width indent
2540 Looks up translation state in all configured nat instances.
2541 If an entry is found, packet is aliased according to that entry.
2542 If no entry was found in any of the instances, packet is passed unchanged,
2543 and no new entry will be created.
2545 .Sx MULTIPLE INSTANCES
2548 for more information.
2550 Uses argument supplied in lookup table.
2553 section below for more information on lookup tables.
2556 To let the packet continue after being (de)aliased, set the sysctl variable
2557 .Va net.inet.ip.fw.one_pass
2559 For more information about aliasing modes, refer to
2563 for some examples about nat usage.
2564 .Ss REDIRECT AND LSNAT SUPPORT IN IPFW
2565 Redirect and LSNAT support follow closely the syntax used in
2569 for some examples on how to do redirect and lsnat.
2570 .Ss SCTP NAT SUPPORT
2571 SCTP nat can be configured in a similar manner to TCP through the
2574 The main difference is that
2576 does not do port translation.
2577 Since the local and global side ports will be the same,
2578 there is no need to specify both.
2579 Ports are redirected as follows:
2580 .Bd -ragged -offset indent
2586 .Cm redirect_port sctp
2587 .Ar ip_address [,addr_list] {[port | port-port] [,ports]}
2593 configuration can be done in real-time through the
2596 All may be changed dynamically, though the hash_table size will only
2601 .Sx SYSCTL VARIABLES
2604 Tunables can be set in
2610 before ipfw module gets loaded.
2611 .Bl -tag -width indent
2612 .It Va net.inet.ip.fw.default_to_accept: No 0
2613 Defines ipfw last rule behavior.
2614 This value overrides
2615 .Cd "options IPFW_DEFAULT_TO_(ACCEPT|DENY)"
2616 from kernel configuration file.
2617 .It Va net.inet.ip.fw.tables_max: No 128
2618 Defines number of tables available in ipfw.
2619 Number cannot exceed 65534.
2621 .Sh SYSCTL VARIABLES
2624 variables controls the behaviour of the firewall and
2626 .Pq Nm dummynet , bridge , sctp nat .
2627 These are shown below together with their default value
2628 (but always check with the
2630 command what value is actually in use) and meaning:
2631 .Bl -tag -width indent
2632 .It Va net.inet.ip.alias.sctp.accept_global_ootb_addip: No 0
2635 responds to receipt of global OOTB ASCONF-AddIP:
2636 .Bl -tag -width indent
2638 No response (unless a partially matching association exists -
2639 ports and vtags match but global address does not)
2642 will accept and process all OOTB global AddIP messages.
2645 Option 1 should never be selected as this forms a security risk.
2647 establish multiple fake associations by sending AddIP messages.
2648 .It Va net.inet.ip.alias.sctp.chunk_proc_limit: No 5
2649 Defines the maximum number of chunks in an SCTP packet that will be
2651 packet that matches an existing association.
2652 This value is enforced to be greater or equal than
2653 .Cm net.inet.ip.alias.sctp.initialising_chunk_proc_limit .
2655 a DoS risk yet setting too low a value may result in
2656 important control chunks in
2657 the packet not being located and parsed.
2658 .It Va net.inet.ip.alias.sctp.error_on_ootb: No 1
2661 responds to any Out-of-the-Blue (OOTB) packets with ErrorM packets.
2662 An OOTB packet is a packet that arrives with no existing association
2665 and is not an INIT or ASCONF-AddIP packet:
2666 .Bl -tag -width indent
2668 ErrorM is never sent in response to OOTB packets.
2670 ErrorM is only sent to OOTB packets received on the local side.
2672 ErrorM is sent to the local side and on the global side ONLY if there is a
2673 partial match (ports and vtags match but the source global IP does not).
2674 This value is only useful if the
2676 is tracking global IP addresses.
2678 ErrorM is sent in response to all OOTB packets on both
2679 the local and global side
2683 At the moment the default is 0, since the ErrorM packet is not yet
2684 supported by most SCTP stacks.
2685 When it is supported, and if not tracking
2686 global addresses, we recommend setting this value to 1 to allow
2687 multi-homed local hosts to function with the
2689 To track global addresses, we recommend setting this value to 2 to
2690 allow global hosts to be informed when they need to (re)send an
2692 Value 3 should never be chosen (except for debugging) as the
2694 will respond to all OOTB global packets (a DoS risk).
2695 .It Va net.inet.ip.alias.sctp.hashtable_size: No 2003
2696 Size of hash tables used for
2698 lookups (100 < prime_number > 1000001).
2701 size for any future created
2703 instance and therefore must be set prior to creating a
2706 The table sizes may be changed to suit specific needs.
2707 If there will be few
2708 concurrent associations, and memory is scarce, you may make these smaller.
2709 If there will be many thousands (or millions) of concurrent associations, you
2710 should make these larger.
2711 A prime number is best for the table size.
2713 update function will adjust your input value to the next highest prime number.
2714 .It Va net.inet.ip.alias.sctp.holddown_time: No 0
2715 Hold association in table for this many seconds after receiving a
2717 This allows endpoints to correct shutdown gracefully if a
2718 shutdown_complete is lost and retransmissions are required.
2719 .It Va net.inet.ip.alias.sctp.init_timer: No 15
2720 Timeout value while waiting for (INIT-ACK|AddIP-ACK).
2721 This value cannot be 0.
2722 .It Va net.inet.ip.alias.sctp.initialising_chunk_proc_limit: No 2
2723 Defines the maximum number of chunks in an SCTP packet that will be parsed when
2724 no existing association exists that matches that packet.
2726 will only be an INIT or ASCONF-AddIP packet.
2727 A higher value may become a DoS
2728 risk as malformed packets can consume processing resources.
2729 .It Va net.inet.ip.alias.sctp.param_proc_limit: No 25
2730 Defines the maximum number of parameters within a chunk that will be
2733 As for other similar sysctl variables, larger values pose a DoS risk.
2734 .It Va net.inet.ip.alias.sctp.log_level: No 0
2735 Level of detail in the system log messages (0 \- minimal, 1 \- event,
2736 2 \- info, 3 \- detail, 4 \- debug, 5 \- max debug).
2738 option in high loss environments.
2739 .It Va net.inet.ip.alias.sctp.shutdown_time: No 15
2740 Timeout value while waiting for SHUTDOWN-COMPLETE.
2741 This value cannot be 0.
2742 .It Va net.inet.ip.alias.sctp.track_global_addresses: No 0
2743 Enables/disables global IP address tracking within the
2746 upper limit on the number of addresses tracked for each association:
2747 .Bl -tag -width indent
2749 Global tracking is disabled
2751 Enables tracking, the maximum number of addresses tracked for each
2752 association is limited to this value
2755 This variable is fully dynamic, the new value will be adopted for all newly
2756 arriving associations, existing associations are treated
2757 as they were previously.
2758 Global tracking will decrease the number of collisions within the
2761 of increased processing load, memory usage, complexity, and possible
2764 problems in complex networks with multiple
2766 We recommend not tracking
2767 global IP addresses, this will still result in a fully functional
2769 .It Va net.inet.ip.alias.sctp.up_timer: No 300
2770 Timeout value to keep an association up with no traffic.
2771 This value cannot be 0.
2772 .It Va net.inet.ip.dummynet.expire : No 1
2773 Lazily delete dynamic pipes/queue once they have no pending traffic.
2774 You can disable this by setting the variable to 0, in which case
2775 the pipes/queues will only be deleted when the threshold is reached.
2776 .It Va net.inet.ip.dummynet.hash_size : No 64
2777 Default size of the hash table used for dynamic pipes/queues.
2778 This value is used when no
2780 option is specified when configuring a pipe/queue.
2781 .It Va net.inet.ip.dummynet.io_fast : No 0
2782 If set to a non-zero value,
2787 operation (see above) is enabled.
2788 .It Va net.inet.ip.dummynet.io_pkt
2789 Number of packets passed to
2791 .It Va net.inet.ip.dummynet.io_pkt_drop
2792 Number of packets dropped by
2794 .It Va net.inet.ip.dummynet.io_pkt_fast
2795 Number of packets bypassed by the
2798 .It Va net.inet.ip.dummynet.max_chain_len : No 16
2799 Target value for the maximum number of pipes/queues in a hash bucket.
2801 .Cm max_chain_len*hash_size
2802 is used to determine the threshold over which empty pipes/queues
2803 will be expired even when
2804 .Cm net.inet.ip.dummynet.expire=0 .
2805 .It Va net.inet.ip.dummynet.red_lookup_depth : No 256
2806 .It Va net.inet.ip.dummynet.red_avg_pkt_size : No 512
2807 .It Va net.inet.ip.dummynet.red_max_pkt_size : No 1500
2808 Parameters used in the computations of the drop probability
2809 for the RED algorithm.
2810 .It Va net.inet.ip.dummynet.pipe_byte_limit : No 1048576
2811 .It Va net.inet.ip.dummynet.pipe_slot_limit : No 100
2812 The maximum queue size that can be specified in bytes or packets.
2813 These limits prevent accidental exhaustion of resources such as mbufs.
2814 If you raise these limits,
2815 you should make sure the system is configured so that sufficient resources
2817 .It Va net.inet.ip.fw.autoinc_step : No 100
2818 Delta between rule numbers when auto-generating them.
2819 The value must be in the range 1..1000.
2820 .It Va net.inet.ip.fw.curr_dyn_buckets : Va net.inet.ip.fw.dyn_buckets
2821 The current number of buckets in the hash table for dynamic rules
2823 .It Va net.inet.ip.fw.debug : No 1
2824 Controls debugging messages produced by
2826 .It Va net.inet.ip.fw.default_rule : No 65535
2827 The default rule number (read-only).
2829 .Nm , the default rule is the last one, so its number
2830 can also serve as the highest number allowed for a rule.
2831 .It Va net.inet.ip.fw.dyn_buckets : No 256
2832 The number of buckets in the hash table for dynamic rules.
2833 Must be a power of 2, up to 65536.
2834 It only takes effect when all dynamic rules have expired, so you
2835 are advised to use a
2837 command to make sure that the hash table is resized.
2838 .It Va net.inet.ip.fw.dyn_count : No 3
2839 Current number of dynamic rules
2841 .It Va net.inet.ip.fw.dyn_keepalive : No 1
2842 Enables generation of keepalive packets for
2844 rules on TCP sessions.
2845 A keepalive is generated to both
2846 sides of the connection every 5 seconds for the last 20
2847 seconds of the lifetime of the rule.
2848 .It Va net.inet.ip.fw.dyn_max : No 8192
2849 Maximum number of dynamic rules.
2850 When you hit this limit, no more dynamic rules can be
2851 installed until old ones expire.
2852 .It Va net.inet.ip.fw.dyn_ack_lifetime : No 300
2853 .It Va net.inet.ip.fw.dyn_syn_lifetime : No 20
2854 .It Va net.inet.ip.fw.dyn_fin_lifetime : No 1
2855 .It Va net.inet.ip.fw.dyn_rst_lifetime : No 1
2856 .It Va net.inet.ip.fw.dyn_udp_lifetime : No 5
2857 .It Va net.inet.ip.fw.dyn_short_lifetime : No 30
2858 These variables control the lifetime, in seconds, of dynamic
2860 Upon the initial SYN exchange the lifetime is kept short,
2861 then increased after both SYN have been seen, then decreased
2862 again during the final FIN exchange or when a RST is received.
2864 .Em dyn_fin_lifetime
2866 .Em dyn_rst_lifetime
2867 must be strictly lower than 5 seconds, the period of
2868 repetition of keepalives.
2869 The firewall enforces that.
2870 .It Va net.inet.ip.fw.enable : No 1
2871 Enables the firewall.
2872 Setting this variable to 0 lets you run your machine without
2873 firewall even if compiled in.
2874 .It Va net.inet6.ip6.fw.enable : No 1
2875 provides the same functionality as above for the IPv6 case.
2876 .It Va net.inet.ip.fw.one_pass : No 1
2877 When set, the packet exiting from the
2881 node is not passed though the firewall again.
2882 Otherwise, after an action, the packet is
2883 reinjected into the firewall at the next rule.
2884 .It Va net.inet.ip.fw.tables_max : No 128
2885 Maximum number of tables.
2886 .It Va net.inet.ip.fw.verbose : No 1
2887 Enables verbose messages.
2888 .It Va net.inet.ip.fw.verbose_limit : No 0
2889 Limits the number of messages produced by a verbose firewall.
2890 .It Va net.inet6.ip6.fw.deny_unknown_exthdrs : No 1
2891 If enabled packets with unknown IPv6 Extension Headers will be denied.
2892 .It Va net.link.ether.ipfw : No 0
2893 Controls whether layer-2 packets are passed to
2896 .It Va net.link.bridge.ipfw : No 0
2897 Controls whether bridged packets are passed to
2902 There are far too many possible uses of
2904 so this Section will only give a small set of examples.
2906 .Ss BASIC PACKET FILTERING
2907 This command adds an entry which denies all tcp packets from
2908 .Em cracker.evil.org
2909 to the telnet port of
2911 from being forwarded by the host:
2913 .Dl "ipfw add deny tcp from cracker.evil.org to wolf.tambov.su telnet"
2915 This one disallows any connection from the entire cracker's
2918 .Dl "ipfw add deny ip from 123.45.67.0/24 to my.host.org"
2920 A first and efficient way to limit access (not using dynamic rules)
2921 is the use of the following rules:
2923 .Dl "ipfw add allow tcp from any to any established"
2924 .Dl "ipfw add allow tcp from net1 portlist1 to net2 portlist2 setup"
2925 .Dl "ipfw add allow tcp from net3 portlist3 to net3 portlist3 setup"
2927 .Dl "ipfw add deny tcp from any to any"
2929 The first rule will be a quick match for normal TCP packets,
2930 but it will not match the initial SYN packet, which will be
2933 rules only for selected source/destination pairs.
2934 All other SYN packets will be rejected by the final
2938 If you administer one or more subnets, you can take advantage
2939 of the address sets and or-blocks and write extremely
2940 compact rulesets which selectively enable services to blocks
2941 of clients, as below:
2943 .Dl "goodguys=\*q{ 10.1.2.0/24{20,35,66,18} or 10.2.3.0/28{6,3,11} }\*q"
2944 .Dl "badguys=\*q10.1.2.0/24{8,38,60}\*q"
2946 .Dl "ipfw add allow ip from ${goodguys} to any"
2947 .Dl "ipfw add deny ip from ${badguys} to any"
2948 .Dl "... normal policies ..."
2952 option could be used to do automated anti-spoofing by adding the
2953 following to the top of a ruleset:
2955 .Dl "ipfw add deny ip from any to any not verrevpath in"
2957 This rule drops all incoming packets that appear to be coming to the
2958 system on the wrong interface.
2959 For example, a packet with a source
2960 address belonging to a host on a protected internal network would be
2961 dropped if it tried to enter the system from an external interface.
2965 option could be used to do similar but more restricted anti-spoofing
2966 by adding the following to the top of a ruleset:
2968 .Dl "ipfw add deny ip from any to any not antispoof in"
2970 This rule drops all incoming packets that appear to be coming from another
2971 directly connected system but on the wrong interface.
2972 For example, a packet with a source address of
2973 .Li 192.168.0.0/24 ,
2980 In order to protect a site from flood attacks involving fake
2981 TCP packets, it is safer to use dynamic rules:
2983 .Dl "ipfw add check-state"
2984 .Dl "ipfw add deny tcp from any to any established"
2985 .Dl "ipfw add allow tcp from my-net to any setup keep-state"
2987 This will let the firewall install dynamic rules only for
2988 those connection which start with a regular SYN packet coming
2989 from the inside of our network.
2990 Dynamic rules are checked when encountering the first
2999 rule should usually be placed near the beginning of the
3000 ruleset to minimize the amount of work scanning the ruleset.
3001 Your mileage may vary.
3003 To limit the number of connections a user can open
3004 you can use the following type of rules:
3006 .Dl "ipfw add allow tcp from my-net/24 to any setup limit src-addr 10"
3007 .Dl "ipfw add allow tcp from any to me setup limit src-addr 4"
3009 The former (assuming it runs on a gateway) will allow each host
3010 on a /24 network to open at most 10 TCP connections.
3011 The latter can be placed on a server to make sure that a single
3012 client does not use more than 4 simultaneous connections.
3015 stateful rules can be subject to denial-of-service attacks
3016 by a SYN-flood which opens a huge number of dynamic rules.
3017 The effects of such attacks can be partially limited by
3020 variables which control the operation of the firewall.
3022 Here is a good usage of the
3024 command to see accounting records and timestamp information:
3028 or in short form without timestamps:
3032 which is equivalent to:
3036 Next rule diverts all incoming packets from 192.168.2.0/24
3037 to divert port 5000:
3039 .Dl ipfw divert 5000 ip from 192.168.2.0/24 to any in
3041 The following rules show some of the applications of
3045 for simulations and the like.
3047 This rule drops random incoming packets with a probability
3050 .Dl "ipfw add prob 0.05 deny ip from any to any in"
3052 A similar effect can be achieved making use of
3056 .Dl "ipfw add pipe 10 ip from any to any"
3057 .Dl "ipfw pipe 10 config plr 0.05"
3059 We can use pipes to artificially limit bandwidth, e.g.\& on a
3060 machine acting as a router, if we want to limit traffic from
3061 local clients on 192.168.2.0/24 we do:
3063 .Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out"
3064 .Dl "ipfw pipe 1 config bw 300Kbit/s queue 50KBytes"
3066 note that we use the
3068 modifier so that the rule is not used twice.
3069 Remember in fact that
3071 rules are checked both on incoming and outgoing packets.
3073 Should we want to simulate a bidirectional link with bandwidth
3074 limitations, the correct way is the following:
3076 .Dl "ipfw add pipe 1 ip from any to any out"
3077 .Dl "ipfw add pipe 2 ip from any to any in"
3078 .Dl "ipfw pipe 1 config bw 64Kbit/s queue 10Kbytes"
3079 .Dl "ipfw pipe 2 config bw 64Kbit/s queue 10Kbytes"
3081 The above can be very useful, e.g.\& if you want to see how
3082 your fancy Web page will look for a residential user who
3083 is connected only through a slow link.
3084 You should not use only one pipe for both directions, unless
3085 you want to simulate a half-duplex medium (e.g.\& AppleTalk,
3087 It is not necessary that both pipes have the same configuration,
3088 so we can also simulate asymmetric links.
3090 Should we want to verify network performance with the RED queue
3091 management algorithm:
3093 .Dl "ipfw add pipe 1 ip from any to any"
3094 .Dl "ipfw pipe 1 config bw 500Kbit/s queue 100 red 0.002/30/80/0.1"
3096 Another typical application of the traffic shaper is to
3097 introduce some delay in the communication.
3098 This can significantly affect applications which do a lot of Remote
3099 Procedure Calls, and where the round-trip-time of the
3100 connection often becomes a limiting factor much more than
3103 .Dl "ipfw add pipe 1 ip from any to any out"
3104 .Dl "ipfw add pipe 2 ip from any to any in"
3105 .Dl "ipfw pipe 1 config delay 250ms bw 1Mbit/s"
3106 .Dl "ipfw pipe 2 config delay 250ms bw 1Mbit/s"
3108 Per-flow queueing can be useful for a variety of purposes.
3109 A very simple one is counting traffic:
3111 .Dl "ipfw add pipe 1 tcp from any to any"
3112 .Dl "ipfw add pipe 1 udp from any to any"
3113 .Dl "ipfw add pipe 1 ip from any to any"
3114 .Dl "ipfw pipe 1 config mask all"
3116 The above set of rules will create queues (and collect
3117 statistics) for all traffic.
3118 Because the pipes have no limitations, the only effect is
3119 collecting statistics.
3120 Note that we need 3 rules, not just the last one, because
3123 tries to match IP packets it will not consider ports, so we
3124 would not see connections on separate ports as different
3127 A more sophisticated example is limiting the outbound traffic
3128 on a net with per-host limits, rather than per-network limits:
3130 .Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out"
3131 .Dl "ipfw add pipe 2 ip from any to 192.168.2.0/24 in"
3132 .Dl "ipfw pipe 1 config mask src-ip 0x000000ff bw 200Kbit/s queue 20Kbytes"
3133 .Dl "ipfw pipe 2 config mask dst-ip 0x000000ff bw 200Kbit/s queue 20Kbytes"
3135 In the following example, we need to create several traffic bandwidth
3136 classes and we need different hosts/networks to fall into different classes.
3137 We create one pipe for each class and configure them accordingly.
3138 Then we create a single table and fill it with IP subnets and addresses.
3139 For each subnet/host we set the argument equal to the number of the pipe
3141 Then we classify traffic using a single rule:
3143 .Dl "ipfw pipe 1 config bw 1000Kbyte/s"
3144 .Dl "ipfw pipe 4 config bw 4000Kbyte/s"
3146 .Dl "ipfw table 1 add 192.168.2.0/24 1"
3147 .Dl "ipfw table 1 add 192.168.0.0/27 4"
3148 .Dl "ipfw table 1 add 192.168.0.2 1"
3150 .Dl "ipfw add pipe tablearg ip from table(1) to any"
3154 action, the table entries may include hostnames and IP addresses.
3156 .Dl "ipfw table 1 add 192.168.2.0/24 10.23.2.1"
3157 .Dl "ipfw table 1 add 192.168.0.0/27 router1.dmz"
3159 .Dl "ipfw add 100 fwd tablearg ip from any to table(1)"
3161 In the following example per-interface firewall is created:
3163 .Dl "ipfw table 10 add vlan20 12000"
3164 .Dl "ipfw table 10 add vlan30 13000"
3165 .Dl "ipfw table 20 add vlan20 22000"
3166 .Dl "ipfw table 20 add vlan30 23000"
3168 .Dl "ipfw add 100 ipfw skipto tablearg ip from any to any recv 'table(10)' in"
3169 .Dl "ipfw add 200 ipfw skipto tablearg ip from any to any xmit 'table(10)' out"
3171 To add a set of rules atomically, e.g.\& set 18:
3173 .Dl "ipfw set disable 18"
3174 .Dl "ipfw add NN set 18 ... # repeat as needed"
3175 .Dl "ipfw set enable 18"
3177 To delete a set of rules atomically the command is simply:
3179 .Dl "ipfw delete set 18"
3181 To test a ruleset and disable it and regain control if something goes wrong:
3183 .Dl "ipfw set disable 18"
3184 .Dl "ipfw add NN set 18 ... # repeat as needed"
3185 .Dl "ipfw set enable 18; echo done; sleep 30 && ipfw set disable 18"
3187 Here if everything goes well, you press control-C before the "sleep"
3188 terminates, and your ruleset will be left active.
3189 Otherwise, e.g.\& if
3190 you cannot access your box, the ruleset will be disabled after
3191 the sleep terminates thus restoring the previous situation.
3193 To show rules of the specific set:
3195 .Dl "ipfw set 18 show"
3197 To show rules of the disabled set:
3199 .Dl "ipfw -S set 18 show"
3201 To clear a specific rule counters of the specific set:
3203 .Dl "ipfw set 18 zero NN"
3205 To delete a specific rule of the specific set:
3207 .Dl "ipfw set 18 delete NN"
3208 .Ss NAT, REDIRECT AND LSNAT
3209 First redirect all the traffic to nat instance 123:
3211 .Dl "ipfw add nat 123 all from any to any"
3213 Then to configure nat instance 123 to alias all the outgoing traffic with ip
3214 192.168.0.123, blocking all incoming connections, trying to keep
3215 same ports on both sides, clearing aliasing table on address change
3216 and keeping a log of traffic/link statistics:
3218 .Dl "ipfw nat 123 config ip 192.168.0.123 log deny_in reset same_ports"
3220 Or to change address of instance 123, aliasing table will be cleared (see
3223 .Dl "ipfw nat 123 config ip 10.0.0.1"
3225 To see configuration of nat instance 123:
3227 .Dl "ipfw nat 123 show config"
3229 To show logs of all the instances in range 111-999:
3231 .Dl "ipfw nat 111-999 show"
3233 To see configurations of all instances:
3235 .Dl "ipfw nat show config"
3237 Or a redirect rule with mixed modes could looks like:
3239 .Dl "ipfw nat 123 config redirect_addr 10.0.0.1 10.0.0.66"
3240 .Dl " redirect_port tcp 192.168.0.1:80 500"
3241 .Dl " redirect_proto udp 192.168.1.43 192.168.1.1"
3242 .Dl " redirect_addr 192.168.0.10,192.168.0.11"
3243 .Dl " 10.0.0.100 # LSNAT"
3244 .Dl " redirect_port tcp 192.168.0.1:80,192.168.0.10:22"
3247 or it could be split in:
3249 .Dl "ipfw nat 1 config redirect_addr 10.0.0.1 10.0.0.66"
3250 .Dl "ipfw nat 2 config redirect_port tcp 192.168.0.1:80 500"
3251 .Dl "ipfw nat 3 config redirect_proto udp 192.168.1.43 192.168.1.1"
3252 .Dl "ipfw nat 4 config redirect_addr 192.168.0.10,192.168.0.11,192.168.0.12"
3254 .Dl "ipfw nat 5 config redirect_port tcp"
3255 .Dl " 192.168.0.1:80,192.168.0.10:22,192.168.0.20:25 500"
3276 utility first appeared in
3281 Stateful extensions were introduced in
3284 was introduced in Summer 2002.
3286 .An Ugen J. S. Antsilevich ,
3287 .An Poul-Henning Kamp ,
3293 API based upon code written by
3297 Dummynet has been introduced by Luigi Rizzo in 1997-1998.
3299 Some early work (1999-2000) on the
3301 traffic shaper supported by Akamba Corp.
3303 The ipfw core (ipfw2) has been completely redesigned and
3304 reimplemented by Luigi Rizzo in summer 2002.
3307 options have been added by various developer over the years.
3310 In-kernel NAT support written by
3311 .An Paolo Pisati Aq piso@FreeBSD.org
3312 as part of a Summer of Code 2005 project.
3316 support has been developed by
3317 .An The Centre for Advanced Internet Architectures (CAIA) Aq http://www.caia.swin.edu.au .
3318 The primary developers and maintainers are David Hayes and Jason But.
3319 For further information visit:
3320 .Aq http://www.caia.swin.edu.au/urp/SONATA
3322 Delay profiles have been developed by Alessandro Cerri and
3323 Luigi Rizzo, supported by the
3324 European Commission within Projects Onelab and Onelab2.
3326 The syntax has grown over the years and sometimes it might be confusing.
3327 Unfortunately, backward compatibility prevents cleaning up mistakes
3328 made in the definition of the syntax.
3332 Misconfiguring the firewall can put your computer in an unusable state,
3333 possibly shutting down network services and requiring console access to
3334 regain control of it.
3336 Incoming packet fragments diverted by
3338 are reassembled before delivery to the socket.
3339 The action used on those packet is the one from the
3340 rule which matches the first fragment of the packet.
3342 Packets diverted to userland, and then reinserted by a userland process
3343 may lose various packet attributes.
3344 The packet source interface name
3345 will be preserved if it is shorter than 8 bytes and the userland process
3346 saves and reuses the sockaddr_in
3349 otherwise, it may be lost.
3350 If a packet is reinserted in this manner, later rules may be incorrectly
3351 applied, making the order of
3353 rules in the rule sequence very important.
3355 Dummynet drops all packets with IPv6 link-local addresses.
3361 may not behave as expected.
3362 In particular, incoming SYN packets may
3363 have no uid or gid associated with them since they do not yet belong
3364 to a TCP connection, and the uid/gid associated with a packet may not
3365 be as expected if the associated process calls
3367 or similar system calls.
3369 Rule syntax is subject to the command line environment and some patterns
3370 may need to be escaped with the backslash character
3371 or quoted appropriately.
3373 Due to the architecture of
3375 ipfw nat is not compatible with the TCP segmentation offloading (TSO).
3376 Thus, to reliably nat your network traffic, please disable TSO
3380 ICMP error messages are not implicitly matched by dynamic rules
3381 for the respective conversations.
3382 To avoid failures of network error detection and path MTU discovery,
3383 ICMP error messages may need to be allowed explicitly through static
3390 actions may lead to confusing behaviour if ruleset has mistakes,
3391 and/or interaction with other subsystems (netgraph, dummynet, etc.) is used.
3392 One possible case for this is packet leaving
3394 in subroutine on the input pass, while later on output encountering unpaired
3397 As the call stack is kept intact after input pass, packet will suddenly
3398 return to the rule number used on input pass, not on output one.
3399 Order of processing should be checked carefully to avoid such mistakes.