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 .Oo Cm set Ar N Oc Cm table Ar name Cm create Ar create-options
53 .Oo Cm set Ar N Oc Cm table
57 .Oo Cm set Ar N Oc Cm table Ar name Cm modify Ar modify-options
59 .Oo Cm set Ar N Oc Cm table Ar name Cm swap Ar name
61 .Oo Cm set Ar N Oc Cm table Ar name Cm add Ar table-key Op Ar value
63 .Oo Cm set Ar N Oc Cm table Ar name Cm add Op Ar table-key Ar value ...
65 .Oo Cm set Ar N Oc Cm table Ar name Cm atomic add Op Ar table-key Ar value ...
67 .Oo Cm set Ar N Oc Cm table Ar name Cm delete Op Ar table-key ...
69 .Oo Cm set Ar N Oc Cm table Ar name Cm lookup Ar addr
71 .Oo Cm set Ar N Oc Cm table Ar name Cm lock
73 .Oo Cm set Ar N Oc Cm table Ar name Cm unlock
75 .Oo Cm set Ar N Oc Cm table
79 .Oo Cm set Ar N Oc Cm table
83 .Oo Cm set Ar N Oc Cm table
87 .Oo Cm set Ar N Oc Cm table
90 .Ss DUMMYNET CONFIGURATION (TRAFFIC SHAPER AND PACKET SCHEDULER)
92 .Brq Cm pipe | queue | sched
98 .Brq Cm pipe | queue | sched
99 .Brq Cm delete | list | show
118 .Ss STATEFUL IPv6/IPv4 NETWORK ADDRESS AND PROTOCOL TRANSLATION
120 .Oo Cm set Ar N Oc Cm nat64lsn Ar name Cm create Ar create-options
122 .Oo Cm set Ar N Oc Cm nat64lsn Ar name Cm config Ar config-options
124 .Oo Cm set Ar N Oc Cm nat64lsn
129 .Oo Cm set Ar N Oc Cm nat64lsn
133 .Oo Cm set Ar N Oc Cm nat64lsn Ar name Cm stats Op Cm reset
134 .Ss STATELESS IPv6/IPv4 NETWORK ADDRESS AND PROTOCOL TRANSLATION
136 .Oo Cm set Ar N Oc Cm nat64stl Ar name Cm create Ar create-options
138 .Oo Cm set Ar N Oc Cm nat64stl Ar name Cm config Ar config-options
140 .Oo Cm set Ar N Oc Cm nat64stl
144 .Oo Cm set Ar N Oc Cm nat64stl
148 .Oo Cm set Ar N Oc Cm nat64stl Ar name Cm stats Op Cm reset
149 .Ss IPv6-to-IPv6 NETWORK PREFIX TRANSLATION
151 .Oo Cm set Ar N Oc Cm nptv6 Ar name Cm create Ar create-options
153 .Oo Cm set Ar N Oc Cm nptv6
157 .Oo Cm set Ar N Oc Cm nptv6
161 .Oo Cm set Ar N Oc Cm nptv6 Ar name Cm stats Op Cm reset
162 .Ss INTERNAL DIAGNOSTICS
172 utility is the user interface for controlling the
176 traffic shaper/packet scheduler, and the
177 in-kernel NAT services.
179 A firewall configuration, or
183 numbered from 1 to 65535.
184 Packets are passed to the firewall
185 from a number of different places in the protocol stack
186 (depending on the source and destination of the packet,
187 it is possible for the firewall to be
188 invoked multiple times on the same packet).
189 The packet passed to the firewall is compared
190 against each of the rules in the
193 (multiple rules with the same number are permitted, in which case
194 they are processed in order of insertion).
195 When a match is found, the action corresponding to the
196 matching rule is performed.
198 Depending on the action and certain system settings, packets
199 can be reinjected into the firewall at some rule after the
200 matching one for further processing.
202 A ruleset always includes a
204 rule (numbered 65535) which cannot be modified or deleted,
205 and matches all packets.
206 The action associated with the
212 depending on how the kernel is configured.
214 If the ruleset includes one or more rules with the
221 the firewall will have a
223 behaviour, i.e., upon a match it will create
225 i.e., rules that match packets with the same 5-tuple
226 (protocol, source and destination addresses and ports)
227 as the packet which caused their creation.
228 Dynamic rules, which have a limited lifetime, are checked
229 at the first occurrence of a
234 rule, and are typically used to open the firewall on-demand to
235 legitimate traffic only.
242 for all packets (not only these matched by the rule) but
249 .Sx STATEFUL FIREWALL
252 Sections below for more information on the stateful behaviour of
255 All rules (including dynamic ones) have a few associated counters:
256 a packet count, a byte count, a log count and a timestamp
257 indicating the time of the last match.
258 Counters can be displayed or reset with
262 Each rule belongs to one of 32 different
266 commands to atomically manipulate sets, such as enable,
267 disable, swap sets, move all rules in a set to another
268 one, delete all rules in a set.
269 These can be useful to
270 install temporary configurations, or to test them.
273 for more information on
276 Rules can be added with the
278 command; deleted individually or in groups with the
280 command, and globally (except those in set 31) with the
282 command; displayed, optionally with the content of the
288 Finally, counters can be reset with the
295 The following general options are available when invoking
297 .Bl -tag -width indent
299 Show counter values when listing rules.
302 command implies this option.
304 Only show the action and the comment, not the body of a rule.
308 When entering or showing rules, print them in compact form,
309 i.e., omitting the "ip from any to any" string
310 when this does not carry any additional information.
312 When listing, show dynamic rules in addition to static ones.
316 is specified, also show expired dynamic rules.
318 Run without prompting for confirmation for commands that can cause problems if misused,
321 If there is no tty associated with the process, this is implied.
324 command with this flag ignores possible errors,
325 i.e., nonexistent rule number.
326 And for batched commands execution continues with the next command.
328 When listing a table (see the
330 section below for more information on lookup tables), format values
332 By default, values are shown as integers.
334 Only check syntax of the command strings, without actually passing
337 Try to resolve addresses and service names in output.
339 Be quiet when executing the
349 This is useful when updating rulesets by executing multiple
353 .Ql sh\ /etc/rc.firewall ) ,
354 or by processing a file with many
356 rules across a remote login session.
357 It also stops a table add or delete
358 from failing if the entry already exists or is not present.
360 The reason why this option may be important is that
361 for some of these actions,
363 may print a message; if the action results in blocking the
364 traffic to the remote client,
365 the remote login session will be closed
366 and the rest of the ruleset will not be processed.
367 Access to the console would then be required to recover.
369 When listing rules, show the
371 each rule belongs to.
372 If this flag is not specified, disabled rules will not be
375 When listing pipes, sort according to one of the four
376 counters (total or current packets or bytes).
378 When listing, show last match timestamp converted with ctime().
380 When listing, show last match timestamp as seconds from the epoch.
381 This form can be more convenient for postprocessing by scripts.
383 .Ss LIST OF RULES AND PREPROCESSING
384 To ease configuration, rules can be put into a file which is
387 as shown in the last synopsis line.
391 The file will be read line by line and applied as arguments to the
395 Optionally, a preprocessor can be specified using
399 is to be piped through.
400 Useful preprocessors include
406 does not start with a slash
408 as its first character, the usual
410 name search is performed.
411 Care should be taken with this in environments where not all
412 file systems are mounted (yet) by the time
414 is being run (e.g.\& when they are mounted over NFS).
417 has been specified, any additional arguments are passed on to the preprocessor
419 This allows for flexible configuration files (like conditionalizing
420 them on the local hostname) and the use of macros to centralize
421 frequently required arguments like IP addresses.
422 .Ss TRAFFIC SHAPER CONFIGURATION
428 commands are used to configure the traffic shaper and packet scheduler.
430 .Sx TRAFFIC SHAPER (DUMMYNET) CONFIGURATION
431 Section below for details.
433 If the world and the kernel get out of sync the
435 ABI may break, preventing you from being able to add any rules.
436 This can adversely affect the booting process.
441 to temporarily disable the firewall to regain access to the network,
442 allowing you to fix the problem.
444 A packet is checked against the active ruleset in multiple places
445 in the protocol stack, under control of several sysctl variables.
446 These places and variables are shown below, and it is important to
447 have this picture in mind in order to design a correct ruleset.
448 .Bd -literal -offset indent
451 +----------->-----------+
453 [ip(6)_input] [ip(6)_output] net.inet(6).ip(6).fw.enable=1
456 [ether_demux] [ether_output_frame] net.link.ether.ipfw=1
458 +-->--[bdg_forward]-->--+ net.link.bridge.ipfw=1
464 times the same packet goes through the firewall can
465 vary between 0 and 4 depending on packet source and
466 destination, and system configuration.
468 Note that as packets flow through the stack, headers can be
469 stripped or added to it, and so they may or may not be available
471 E.g., incoming packets will include the MAC header when
475 but the same packets will have the MAC header stripped off when
482 Also note that each packet is always checked against the complete ruleset,
483 irrespective of the place where the check occurs, or the source of the packet.
484 If a rule contains some match patterns or actions which are not valid
485 for the place of invocation (e.g.\& trying to match a MAC header within
489 the match pattern will not match, but a
491 operator in front of such patterns
495 match on those packets.
496 It is thus the responsibility of
497 the programmer, if necessary, to write a suitable ruleset to
498 differentiate among the possible places.
500 rules can be useful here, as an example:
501 .Bd -literal -offset indent
502 # packets from ether_demux or bdg_forward
503 ipfw add 10 skipto 1000 all from any to any layer2 in
504 # packets from ip_input
505 ipfw add 10 skipto 2000 all from any to any not layer2 in
506 # packets from ip_output
507 ipfw add 10 skipto 3000 all from any to any not layer2 out
508 # packets from ether_output_frame
509 ipfw add 10 skipto 4000 all from any to any layer2 out
512 (yes, at the moment there is no way to differentiate between
513 ether_demux and bdg_forward).
515 In general, each keyword or argument must be provided as
516 a separate command line argument, with no leading or trailing
518 Keywords are case-sensitive, whereas arguments may
519 or may not be case-sensitive depending on their nature
520 (e.g.\& uid's are, hostnames are not).
522 Some arguments (e.g., port or address lists) are comma-separated
524 In this case, spaces after commas ',' are allowed to make
525 the line more readable.
526 You can also put the entire
527 command (including flags) into a single argument.
528 E.g., the following forms are equivalent:
529 .Bd -literal -offset indent
530 ipfw -q add deny src-ip 10.0.0.0/24,127.0.0.1/8
531 ipfw -q add deny src-ip 10.0.0.0/24, 127.0.0.1/8
532 ipfw "-q add deny src-ip 10.0.0.0/24, 127.0.0.1/8"
535 The format of firewall rules is the following:
536 .Bd -ragged -offset indent
539 .Op Cm set Ar set_number
540 .Op Cm prob Ar match_probability
542 .Op Cm log Op Cm logamount Ar number
552 where the body of the rule specifies which information is used
553 for filtering packets, among the following:
555 .Bl -tag -width "Source and dest. addresses and ports" -offset XXX -compact
556 .It Layer-2 header fields
558 .It IPv4 and IPv6 Protocol
559 SCTP, TCP, UDP, ICMP, etc.
560 .It Source and dest. addresses and ports
564 .It Transmit and receive interface
566 .It Misc. IP header fields
567 Version, type of service, datagram length, identification,
568 fragment flag (non-zero IP offset),
571 .It IPv6 Extension headers
572 Fragmentation, Hop-by-Hop options,
573 Routing Headers, Source routing rthdr0, Mobile IPv6 rthdr2, IPSec options.
575 .It Misc. TCP header fields
576 TCP flags (SYN, FIN, ACK, RST, etc.),
577 sequence number, acknowledgment number,
585 When the packet can be associated with a local socket.
587 Whether a packet came from a divert socket (e.g.,
589 .It Fib annotation state
590 Whether a packet has been tagged for using a specific FIB (routing table)
591 in future forwarding decisions.
594 Note that some of the above information, e.g.\& source MAC or IP addresses and
595 TCP/UDP ports, can be easily spoofed, so filtering on those fields
596 alone might not guarantee the desired results.
597 .Bl -tag -width indent
599 Each rule is associated with a
601 in the range 1..65535, with the latter reserved for the
604 Rules are checked sequentially by rule number.
605 Multiple rules can have the same number, in which case they are
606 checked (and listed) according to the order in which they have
608 If a rule is entered without specifying a number, the kernel will
609 assign one in such a way that the rule becomes the last one
613 Automatic rule numbers are assigned by incrementing the last
614 non-default rule number by the value of the sysctl variable
615 .Ar net.inet.ip.fw.autoinc_step
616 which defaults to 100.
617 If this is not possible (e.g.\& because we would go beyond the
618 maximum allowed rule number), the number of the last
619 non-default value is used instead.
620 .It Cm set Ar set_number
621 Each rule is associated with a
624 Sets can be individually disabled and enabled, so this parameter
625 is of fundamental importance for atomic ruleset manipulation.
626 It can be also used to simplify deletion of groups of rules.
627 If a rule is entered without specifying a set number,
630 Set 31 is special in that it cannot be disabled,
631 and rules in set 31 are not deleted by the
633 command (but you can delete them with the
634 .Nm ipfw delete set 31
636 Set 31 is also used for the
639 .It Cm prob Ar match_probability
640 A match is only declared with the specified probability
641 (floating point number between 0 and 1).
642 This can be useful for a number of applications such as
643 random packet drop or
646 to simulate the effect of multiple paths leading to out-of-order
649 Note: this condition is checked before any other condition, including
656 .It Cm log Op Cm logamount Ar number
657 Packets matching a rule with the
659 keyword will be made available for logging in two ways:
660 if the sysctl variable
661 .Va net.inet.ip.fw.verbose
662 is set to 0 (default), one can use
667 This pseudo interface can be created after a boot
668 manually by using the following command:
669 .Bd -literal -offset indent
670 # ifconfig ipfw0 create
673 Or, automatically at boot time by adding the following
677 .Bd -literal -offset indent
681 There is no overhead if no
683 is attached to the pseudo interface.
686 .Va net.inet.ip.fw.verbose
687 is set to 1, packets will be logged to
691 facility up to a maximum of
696 is specified, the limit is taken from the sysctl variable
697 .Va net.inet.ip.fw.verbose_limit .
698 In both cases, a value of 0 means unlimited logging.
700 Once the limit is reached, logging can be re-enabled by
701 clearing the logging counter or the packet counter for that entry, see the
705 Note: logging is done after all other packet matching conditions
706 have been successfully verified, and before performing the final
707 action (accept, deny, etc.) on the packet.
709 When a packet matches a rule with the
711 keyword, the numeric tag for the given
713 in the range 1..65534 will be attached to the packet.
714 The tag acts as an internal marker (it is not sent out over
715 the wire) that can be used to identify these packets later on.
716 This can be used, for example, to provide trust between interfaces
717 and to start doing policy-based filtering.
718 A packet can have multiple tags at the same time.
719 Tags are "sticky", meaning once a tag is applied to a packet by a
720 matching rule it exists until explicit removal.
721 Tags are kept with the packet everywhere within the kernel, but are
722 lost when packet leaves the kernel, for example, on transmitting
723 packet out to the network or sending packet to a
727 To check for previously applied tags, use the
730 To delete previously applied tag, use the
734 Note: since tags are kept with the packet everywhere in kernelspace,
735 they can be set and unset anywhere in the kernel network subsystem
738 facility), not only by means of the
744 For example, there can be a specialized
746 node doing traffic analyzing and tagging for later inspecting
748 .It Cm untag Ar number
749 When a packet matches a rule with the
751 keyword, the tag with the number
753 is searched among the tags attached to this packet and,
754 if found, removed from it.
755 Other tags bound to packet, if present, are left untouched.
757 When a packet matches a rule with the
759 keyword, the ALTQ identifier for the given
764 Note that this ALTQ tag is only meaningful for packets going "out" of IPFW,
765 and not being rejected or going to divert sockets.
766 Note that if there is insufficient memory at the time the packet is
767 processed, it will not be tagged, so it is wise to make your ALTQ
768 "default" queue policy account for this.
771 rules match a single packet, only the first one adds the ALTQ classification
773 In doing so, traffic may be shaped by using
774 .Cm count Cm altq Ar queue
775 rules for classification early in the ruleset, then later applying
776 the filtering decision.
781 rules may come later and provide the actual filtering decisions in
782 addition to the fallback ALTQ tag.
786 to set up the queues before IPFW will be able to look them up by name,
787 and if the ALTQ disciplines are rearranged, the rules in containing the
788 queue identifiers in the kernel will likely have gone stale and need
790 Stale queue identifiers will probably result in misclassification.
792 All system ALTQ processing can be turned on or off via
797 .Cm disable Ar altq .
799 .Va net.inet.ip.fw.one_pass
800 is irrelevant to ALTQ traffic shaping, as the actual rule action is followed
801 always after adding an ALTQ tag.
804 A rule can be associated with one of the following actions, which
805 will be executed when the packet matches the body of the rule.
806 .Bl -tag -width indent
807 .It Cm allow | accept | pass | permit
808 Allow packets that match rule.
809 The search terminates.
810 .It Cm check-state Op Ar :flowname | Cm :any
811 Checks the packet against the dynamic ruleset.
812 If a match is found, execute the action associated with
813 the rule which generated this dynamic rule, otherwise
814 move to the next rule.
817 rules do not have a body.
820 rule is found, the dynamic ruleset is checked at the first
827 is symbolic name assigned to dynamic rule by
832 can be used to ignore states flowname when matching.
835 keyword is special name used for compatibility with old rulesets.
837 Update counters for all packets that match rule.
838 The search continues with the next rule.
840 Discard packets that match this rule.
841 The search terminates.
842 .It Cm divert Ar port
843 Divert packets that match this rule to the
847 The search terminates.
848 .It Cm fwd | forward Ar ipaddr | tablearg Ns Op , Ns Ar port
849 Change the next-hop on matching packets to
851 which can be an IP address or a host name.
852 The next hop can also be supplied by the last table
853 looked up for the packet by using the
855 keyword instead of an explicit address.
856 The search terminates if this rule matches.
860 is a local address, then matching packets will be forwarded to
862 (or the port number in the packet if one is not specified in the rule)
863 on the local machine.
867 is not a local address, then the port number
868 (if specified) is ignored, and the packet will be
869 forwarded to the remote address, using the route as found in
870 the local routing table for that IP.
874 rule will not match layer-2 packets (those received
875 on ether_input, ether_output, or bridged).
879 action does not change the contents of the packet at all.
880 In particular, the destination address remains unmodified, so
881 packets forwarded to another system will usually be rejected by that system
882 unless there is a matching rule on that system to capture them.
883 For packets forwarded locally,
884 the local address of the socket will be
885 set to the original destination address of the packet.
888 entry look rather weird but is intended for
889 use with transparent proxy servers.
890 .It Cm nat Ar nat_nr | tablearg
893 (for network address translation, address redirect, etc.):
895 .Sx NETWORK ADDRESS TRANSLATION (NAT)
896 Section for further information.
897 .It Cm nat64lsn Ar name
898 Pass packet to a stateful NAT64 instance (for IPv6/IPv4 network address and
899 protocol translation): see the
900 .Sx IPv6/IPv4 NETWORK ADDRESS AND PROTOCOL TRANSLATION
901 Section for further information.
902 .It Cm nat64stl Ar name
903 Pass packet to a stateless NAT64 instance (for IPv6/IPv4 network address and
904 protocol translation): see the
905 .Sx IPv6/IPv4 NETWORK ADDRESS AND PROTOCOL TRANSLATION
906 Section for further information.
908 Pass packet to a NPTv6 instance (for IPv6-to-IPv6 network prefix translation):
910 .Sx IPv6-to-IPv6 NETWORK PREFIX TRANSLATION (NPTv6)
911 Section for further information.
912 .It Cm pipe Ar pipe_nr
916 (for bandwidth limitation, delay, etc.).
918 .Sx TRAFFIC SHAPER (DUMMYNET) CONFIGURATION
919 Section for further information.
920 The search terminates; however, on exit from the pipe and if
924 .Va net.inet.ip.fw.one_pass
925 is not set, the packet is passed again to the firewall code
926 starting from the next rule.
927 .It Cm queue Ar queue_nr
931 (for bandwidth limitation using WF2Q+).
937 Discard packets that match this rule, and if the
938 packet is a TCP packet, try to send a TCP reset (RST) notice.
939 The search terminates.
941 Discard packets that match this rule, and if the
942 packet is a TCP packet, try to send a TCP reset (RST) notice.
943 The search terminates.
944 .It Cm skipto Ar number | tablearg
945 Skip all subsequent rules numbered less than
947 The search continues with the first rule numbered
950 It is possible to use the
952 keyword with a skipto for a
954 skipto. Skipto may work either in O(log(N)) or in O(1) depending
955 on amount of memory and/or sysctl variables.
958 section for more details.
959 .It Cm call Ar number | tablearg
960 The current rule number is saved in the internal stack and
961 ruleset processing continues with the first rule numbered
964 If later a rule with the
966 action is encountered, the processing returns to the first rule
969 rule plus one or higher
970 (the same behaviour as with packets returning from
975 This could be used to make somewhat like an assembly language
977 calls to rules with common checks for different interfaces, etc.
979 Rule with any number could be called, not just forward jumps as with
981 So, to prevent endless loops in case of mistakes, both
985 actions don't do any jumps and simply go to the next rule if memory
986 cannot be allocated or stack overflowed/underflowed.
988 Internally stack for rule numbers is implemented using
990 facility and currently has size of 16 entries.
991 As mbuf tags are lost when packet leaves the kernel,
993 should not be used in subroutines to avoid endless loops
994 and other undesired effects.
996 Takes rule number saved to internal stack by the last
998 action and returns ruleset processing to the first rule
999 with number greater than number of corresponding
1002 See description of the
1004 action for more details.
1010 and thus are unconditional, but
1012 command-line utility currently requires every action except
1015 While it is sometimes useful to return only on some packets,
1016 usually you want to print just
1019 A workaround for this is to use new syntax and
1022 .Bd -literal -offset indent
1023 # Add a rule without actual body
1024 ipfw add 2999 return via any
1026 # List rules without "from any to any" part
1030 This cosmetic annoyance may be fixed in future releases.
1032 Send a copy of packets matching this rule to the
1034 socket bound to port
1036 The search continues with the next rule.
1037 .It Cm unreach Ar code
1038 Discard packets that match this rule, and try to send an ICMP
1039 unreachable notice with code
1043 is a number from 0 to 255, or one of these aliases:
1044 .Cm net , host , protocol , port ,
1045 .Cm needfrag , srcfail , net-unknown , host-unknown ,
1046 .Cm isolated , net-prohib , host-prohib , tosnet ,
1047 .Cm toshost , filter-prohib , host-precedence
1049 .Cm precedence-cutoff .
1050 The search terminates.
1051 .It Cm unreach6 Ar code
1052 Discard packets that match this rule, and try to send an ICMPv6
1053 unreachable notice with code
1057 is a number from 0, 1, 3 or 4, or one of these aliases:
1058 .Cm no-route, admin-prohib, address
1061 The search terminates.
1062 .It Cm netgraph Ar cookie
1063 Divert packet into netgraph with given
1065 The search terminates.
1066 If packet is later returned from netgraph it is either
1067 accepted or continues with the next rule, depending on
1068 .Va net.inet.ip.fw.one_pass
1070 .It Cm ngtee Ar cookie
1071 A copy of packet is diverted into netgraph, original
1072 packet continues with the next rule.
1075 for more information on
1080 .It Cm setfib Ar fibnum | tablearg
1081 The packet is tagged so as to use the FIB (routing table)
1083 in any subsequent forwarding decisions.
1084 In the current implementation, this is limited to the values 0 through 15, see
1086 Processing continues at the next rule.
1087 It is possible to use the
1089 keyword with setfib.
1090 If the tablearg value is not within the compiled range of fibs,
1091 the packet's fib is set to 0.
1092 .It Cm setdscp Ar DSCP | number | tablearg
1093 Set specified DiffServ codepoint for an IPv4/IPv6 packet.
1094 Processing continues at the next rule.
1095 Supported values are:
1141 Additionally, DSCP value can be specified by number (0..64).
1142 It is also possible to use the
1144 keyword with setdscp.
1145 If the tablearg value is not within the 0..64 range, lower 6 bits of supplied
1147 .It Cm tcp-setmss Ar mss
1148 Set the Maximum Segment Size (MSS) in the TCP segment to value
1152 should be loaded or kernel should have
1153 .Cm options IPFIREWALL_PMOD
1154 to be able use this action.
1155 This command does not change a packet if original MSS value is lower than
1157 Both TCP over IPv4 and over IPv6 are supported.
1158 Regardless of matched a packet or not by the
1160 rule, the search continues with the next rule.
1162 Queue and reassemble IPv4 fragments.
1163 If the packet is not fragmented, counters are updated and
1164 processing continues with the next rule.
1165 If the packet is the last logical fragment, the packet is reassembled and, if
1166 .Va net.inet.ip.fw.one_pass
1167 is set to 0, processing continues with the next rule.
1168 Otherwise, the packet is allowed to pass and the search terminates.
1169 If the packet is a fragment in the middle of a logical group of fragments,
1171 processing stops immediately.
1173 Fragment handling can be tuned via
1174 .Va net.inet.ip.maxfragpackets
1176 .Va net.inet.ip.maxfragsperpacket
1177 which limit, respectively, the maximum number of processable
1178 fragments (default: 800) and
1179 the maximum number of fragments per packet (default: 16).
1181 NOTA BENE: since fragments do not contain port numbers,
1182 they should be avoided with the
1185 Alternatively, direction-based (like
1189 ) and source-based (like
1191 ) match patterns can be used to select fragments.
1193 Usually a simple rule like:
1194 .Bd -literal -offset indent
1195 # reassemble incoming fragments
1196 ipfw add reass all from any to any in
1199 is all you need at the beginning of your ruleset.
1201 Discard packets that match this rule, and if the packet is an SCTP packet,
1202 try to send an SCTP packet containing an ABORT chunk.
1203 The search terminates.
1205 Discard packets that match this rule, and if the packet is an SCTP packet,
1206 try to send an SCTP packet containing an ABORT chunk.
1207 The search terminates.
1210 The body of a rule contains zero or more patterns (such as
1211 specific source and destination addresses or ports,
1212 protocol options, incoming or outgoing interfaces, etc.)
1213 that the packet must match in order to be recognised.
1214 In general, the patterns are connected by (implicit)
1216 operators -- i.e., all must match in order for the
1218 Individual patterns can be prefixed by the
1220 operator to reverse the result of the match, as in
1222 .Dl "ipfw add 100 allow ip from not 1.2.3.4 to any"
1224 Additionally, sets of alternative match patterns
1226 can be constructed by putting the patterns in
1227 lists enclosed between parentheses ( ) or braces { }, and
1230 operator as follows:
1232 .Dl "ipfw add 100 allow ip from { x or not y or z } to any"
1234 Only one level of parentheses is allowed.
1235 Beware that most shells have special meanings for parentheses
1236 or braces, so it is advisable to put a backslash \\ in front of them
1237 to prevent such interpretations.
1239 The body of a rule must in general include a source and destination
1243 can be used in various places to specify that the content of
1244 a required field is irrelevant.
1246 The rule body has the following format:
1247 .Bd -ragged -offset indent
1248 .Op Ar proto Cm from Ar src Cm to Ar dst
1252 The first part (proto from src to dst) is for backward
1253 compatibility with earlier versions of
1257 any match pattern (including MAC headers, IP protocols,
1258 addresses and ports) can be specified in the
1262 Rule fields have the following meaning:
1263 .Bl -tag -width indent
1264 .It Ar proto : protocol | Cm { Ar protocol Cm or ... }
1265 .It Ar protocol : Oo Cm not Oc Ar protocol-name | protocol-number
1266 An IP protocol specified by number or name
1267 (for a complete list see
1268 .Pa /etc/protocols ) ,
1269 or one of the following keywords:
1270 .Bl -tag -width indent
1272 Matches IPv4 packets.
1274 Matches IPv6 packets.
1283 option will be treated as inner protocol.
1291 .Cm { Ar protocol Cm or ... }
1294 is provided for convenience only but its use is deprecated.
1295 .It Ar src No and Ar dst : Bro Cm addr | Cm { Ar addr Cm or ... } Brc Op Oo Cm not Oc Ar ports
1296 An address (or a list, see below)
1297 optionally followed by
1303 with multiple addresses) is provided for convenience only and
1304 its use is discouraged.
1305 .It Ar addr : Oo Cm not Oc Bro
1306 .Cm any | me | me6 |
1307 .Cm table Ns Pq Ar name Ns Op , Ns Ar value
1308 .Ar | addr-list | addr-set
1310 .Bl -tag -width indent
1312 matches any IP address.
1314 matches any IP address configured on an interface in the system.
1316 matches any IPv6 address configured on an interface in the system.
1317 The address list is evaluated at the time the packet is
1319 .It Cm table Ns Pq Ar name Ns Op , Ns Ar value
1320 Matches any IPv4 or IPv6 address for which an entry exists in the lookup table
1322 If an optional 32-bit unsigned
1324 is also specified, an entry will match only if it has this value.
1327 section below for more information on lookup tables.
1329 .It Ar addr-list : ip-addr Ns Op Ns , Ns Ar addr-list
1331 A host or subnet address specified in one of the following ways:
1332 .Bl -tag -width indent
1333 .It Ar numeric-ip | hostname
1334 Matches a single IPv4 address, specified as dotted-quad or a hostname.
1335 Hostnames are resolved at the time the rule is added to the firewall list.
1336 .It Ar addr Ns / Ns Ar masklen
1337 Matches all addresses with base
1339 (specified as an IP address, a network number, or a hostname)
1343 As an example, 1.2.3.4/25 or 1.2.3.0/25 will match
1344 all IP numbers from 1.2.3.0 to 1.2.3.127 .
1345 .It Ar addr Ns : Ns Ar mask
1346 Matches all addresses with base
1348 (specified as an IP address, a network number, or a hostname)
1351 specified as a dotted quad.
1352 As an example, 1.2.3.4:255.0.255.0 or 1.0.3.0:255.0.255.0 will match
1354 This form is advised only for non-contiguous
1356 It is better to resort to the
1357 .Ar addr Ns / Ns Ar masklen
1358 format for contiguous masks, which is more compact and less
1361 .It Ar addr-set : addr Ns Oo Ns / Ns Ar masklen Oc Ns Cm { Ns Ar list Ns Cm }
1362 .It Ar list : Bro Ar num | num-num Brc Ns Op Ns , Ns Ar list
1363 Matches all addresses with base address
1365 (specified as an IP address, a network number, or a hostname)
1366 and whose last byte is in the list between braces { } .
1367 Note that there must be no spaces between braces and
1368 numbers (spaces after commas are allowed).
1369 Elements of the list can be specified as single entries
1373 field is used to limit the size of the set of addresses,
1374 and can have any value between 24 and 32.
1376 it will be assumed as 24.
1378 This format is particularly useful to handle sparse address sets
1379 within a single rule.
1380 Because the matching occurs using a
1381 bitmask, it takes constant time and dramatically reduces
1382 the complexity of rulesets.
1384 As an example, an address specified as 1.2.3.4/24{128,35-55,89}
1385 or 1.2.3.0/24{128,35-55,89}
1386 will match the following IP addresses:
1388 1.2.3.128, 1.2.3.35 to 1.2.3.55, 1.2.3.89 .
1389 .It Ar addr6-list : ip6-addr Ns Op Ns , Ns Ar addr6-list
1391 A host or subnet specified one of the following ways:
1392 .Bl -tag -width indent
1393 .It Ar numeric-ip | hostname
1394 Matches a single IPv6 address as allowed by
1397 Hostnames are resolved at the time the rule is added to the firewall
1399 .It Ar addr Ns / Ns Ar masklen
1400 Matches all IPv6 addresses with base
1402 (specified as allowed by
1408 .It Ar addr Ns / Ns Ar mask
1409 Matches all IPv6 addresses with base
1411 (specified as allowed by
1416 specified as allowed by
1418 As an example, fe::640:0:0/ffff::ffff:ffff:0:0 will match
1420 This form is advised only for non-contiguous
1422 It is better to resort to the
1423 .Ar addr Ns / Ns Ar masklen
1424 format for contiguous masks, which is more compact and less
1428 No support for sets of IPv6 addresses is provided because IPv6 addresses
1429 are typically random past the initial prefix.
1430 .It Ar ports : Bro Ar port | port Ns \&- Ns Ar port Ns Brc Ns Op , Ns Ar ports
1431 For protocols which support port numbers (such as SCTP, TCP and UDP), optional
1433 may be specified as one or more ports or port ranges, separated
1434 by commas but no spaces, and an optional
1439 notation specifies a range of ports (including boundaries).
1443 may be used instead of numeric port values.
1444 The length of the port list is limited to 30 ports or ranges,
1445 though one can specify larger ranges by using an
1449 section of the rule.
1453 can be used to escape the dash
1455 character in a service name (from a shell, the backslash must be
1456 typed twice to avoid the shell itself interpreting it as an escape
1459 .Dl "ipfw add count tcp from any ftp\e\e-data-ftp to any"
1461 Fragmented packets which have a non-zero offset (i.e., not the first
1462 fragment) will never match a rule which has one or more port
1466 option for details on matching fragmented packets.
1468 .Ss RULE OPTIONS (MATCH PATTERNS)
1469 Additional match patterns can be used within
1471 Zero or more of these so-called
1473 can be present in a rule, optionally prefixed by the
1475 operand, and possibly grouped into
1478 The following match patterns can be used (listed in alphabetical order):
1479 .Bl -tag -width indent
1480 .It Cm // this is a comment.
1481 Inserts the specified text as a comment in the rule.
1482 Everything following // is considered as a comment and stored in the rule.
1483 You can have comment-only rules, which are listed as having a
1485 action followed by the comment.
1489 .It Cm defer-immediate-action | defer-action
1490 A rule with this option will not perform normal action
1491 upon a match. This option is intended to be used with
1495 as the dynamic rule, created but ignored on match, will work
1500 .Cm defer-immediate-action
1501 create a dynamic rule and continue with the next rule without actually
1502 performing the action part of this rule. When the rule is later activated
1503 via the state table, the action is performed as usual.
1505 Matches only packets generated by a divert socket.
1506 .It Cm diverted-loopback
1507 Matches only packets coming from a divert socket back into the IP stack
1509 .It Cm diverted-output
1510 Matches only packets going from a divert socket back outward to the IP
1511 stack output for delivery.
1512 .It Cm dst-ip Ar ip-address
1513 Matches IPv4 packets whose destination IP is one of the address(es)
1514 specified as argument.
1515 .It Bro Cm dst-ip6 | dst-ipv6 Brc Ar ip6-address
1516 Matches IPv6 packets whose destination IP is one of the address(es)
1517 specified as argument.
1518 .It Cm dst-port Ar ports
1519 Matches IP packets whose destination port is one of the port(s)
1520 specified as argument.
1522 Matches TCP packets that have the RST or ACK bits set.
1523 .It Cm ext6hdr Ar header
1524 Matches IPv6 packets containing the extended header given by
1526 Supported headers are:
1532 any type of Routing Header
1534 Source routing Routing Header Type 0
1536 Mobile IPv6 Routing Header Type 2
1540 IPSec authentication headers
1542 and IPsec encapsulated security payload headers
1544 .It Cm fib Ar fibnum
1545 Matches a packet that has been tagged to use
1546 the given FIB (routing table) number.
1547 .It Cm flow Ar table Ns Pq Ar name Ns Op , Ns Ar value
1548 Search for the flow entry in lookup table
1550 If not found, the match fails.
1551 Otherwise, the match succeeds and
1553 is set to the value extracted from the table.
1555 This option can be useful to quickly dispatch traffic based on
1556 certain packet fields.
1559 section below for more information on lookup tables.
1560 .It Cm flow-id Ar labels
1561 Matches IPv6 packets containing any of the flow labels given in
1564 is a comma separated list of numeric flow labels.
1566 Matches packets that are fragments and not the first
1567 fragment of an IP datagram.
1568 Note that these packets will not have
1569 the next protocol header (e.g.\& TCP, UDP) so options that look into
1570 these headers cannot match.
1572 Matches all TCP or UDP packets sent by or received for a
1576 may be specified by name or number.
1578 Matches all TCP or UDP packets sent by or received for the
1579 jail whose ID or name is
1581 .It Cm icmptypes Ar types
1582 Matches ICMP packets whose ICMP type is in the list
1584 The list may be specified as any combination of
1585 individual types (numeric) separated by commas.
1586 .Em Ranges are not allowed .
1587 The supported ICMP types are:
1591 destination unreachable
1599 router advertisement
1603 time-to-live exceeded
1615 address mask request
1617 and address mask reply
1619 .It Cm icmp6types Ar types
1620 Matches ICMP6 packets whose ICMP6 type is in the list of
1622 The list may be specified as any combination of
1623 individual types (numeric) separated by commas.
1624 .Em Ranges are not allowed .
1626 Matches incoming or outgoing packets, respectively.
1630 are mutually exclusive (in fact,
1634 .It Cm ipid Ar id-list
1635 Matches IPv4 packets whose
1637 field has value included in
1639 which is either a single value or a list of values or ranges
1640 specified in the same way as
1642 .It Cm iplen Ar len-list
1643 Matches IP packets whose total length, including header and data, is
1646 which is either a single value or a list of values or ranges
1647 specified in the same way as
1649 .It Cm ipoptions Ar spec
1650 Matches packets whose IPv4 header contains the comma separated list of
1651 options specified in
1653 The supported IP options are:
1656 (strict source route),
1658 (loose source route),
1660 (record packet route) and
1663 The absence of a particular option may be denoted
1666 .It Cm ipprecedence Ar precedence
1667 Matches IPv4 packets whose precedence field is equal to
1670 Matches packets that have IPSEC history associated with them
1671 (i.e., the packet comes encapsulated in IPSEC, the kernel
1672 has IPSEC support, and can correctly decapsulate it).
1674 Note that specifying
1676 is different from specifying
1678 as the latter will only look at the specific IP protocol field,
1679 irrespective of IPSEC kernel support and the validity of the IPSEC data.
1681 Further note that this flag is silently ignored in kernels without
1683 It does not affect rule processing when given and the
1684 rules are handled as if with no
1687 .It Cm iptos Ar spec
1688 Matches IPv4 packets whose
1690 field contains the comma separated list of
1691 service types specified in
1693 The supported IP types of service are:
1696 .Pq Dv IPTOS_LOWDELAY ,
1698 .Pq Dv IPTOS_THROUGHPUT ,
1700 .Pq Dv IPTOS_RELIABILITY ,
1702 .Pq Dv IPTOS_MINCOST ,
1704 .Pq Dv IPTOS_ECN_CE .
1705 The absence of a particular type may be denoted
1708 .It Cm dscp spec Ns Op , Ns Ar spec
1709 Matches IPv4/IPv6 packets whose
1711 field value is contained in
1714 Multiple values can be specified via
1715 the comma separated list.
1716 Value can be one of keywords used in
1718 action or exact number.
1719 .It Cm ipttl Ar ttl-list
1720 Matches IPv4 packets whose time to live is included in
1722 which is either a single value or a list of values or ranges
1723 specified in the same way as
1725 .It Cm ipversion Ar ver
1726 Matches IP packets whose IP version field is
1728 .It Cm keep-state Op Ar :flowname
1729 Upon a match, the firewall will create a dynamic rule, whose
1730 default behaviour is to match bidirectional traffic between
1731 source and destination IP/port using the same protocol.
1732 The rule has a limited lifetime (controlled by a set of
1734 variables), and the lifetime is refreshed every time a matching
1738 is used to assign additional to addresses, ports and protocol parameter
1739 to dynamic rule. It can be used for more accurate matching by
1744 keyword is special name used for compatibility with old rulesets.
1746 Matches only layer2 packets, i.e., those passed to
1748 from ether_demux() and ether_output_frame().
1749 .It Cm limit Bro Cm src-addr | src-port | dst-addr | dst-port Brc Ar N Op Ar :flowname
1750 The firewall will only allow
1752 connections with the same
1753 set of parameters as specified in the rule.
1755 of source and destination addresses and ports can be
1757 .It Cm lookup Bro Cm dst-ip | dst-port | src-ip | src-port | uid | jail Brc Ar name
1758 Search an entry in lookup table
1760 that matches the field specified as argument.
1761 If not found, the match fails.
1762 Otherwise, the match succeeds and
1764 is set to the value extracted from the table.
1766 This option can be useful to quickly dispatch traffic based on
1767 certain packet fields.
1770 section below for more information on lookup tables.
1771 .It Cm { MAC | mac } Ar dst-mac src-mac
1772 Match packets with a given
1776 addresses, specified as the
1778 keyword (matching any MAC address), or six groups of hex digits
1779 separated by colons,
1780 and optionally followed by a mask indicating the significant bits.
1781 The mask may be specified using either of the following methods:
1782 .Bl -enum -width indent
1786 followed by the number of significant bits.
1787 For example, an address with 33 significant bits could be specified as:
1789 .Dl "MAC 10:20:30:40:50:60/33 any"
1793 followed by a bitmask specified as six groups of hex digits separated
1795 For example, an address in which the last 16 bits are significant could
1798 .Dl "MAC 10:20:30:40:50:60&00:00:00:00:ff:ff any"
1800 Note that the ampersand character has a special meaning in many shells
1801 and should generally be escaped.
1803 Note that the order of MAC addresses (destination first,
1805 the same as on the wire, but the opposite of the one used for
1807 .It Cm mac-type Ar mac-type
1808 Matches packets whose Ethernet Type field
1809 corresponds to one of those specified as argument.
1811 is specified in the same way as
1813 (i.e., one or more comma-separated single values or ranges).
1814 You can use symbolic names for known values such as
1815 .Em vlan , ipv4, ipv6 .
1816 Values can be entered as decimal or hexadecimal (if prefixed by 0x),
1817 and they are always printed as hexadecimal (unless the
1819 option is used, in which case symbolic resolution will be attempted).
1820 .It Cm proto Ar protocol
1821 Matches packets with the corresponding IP protocol.
1823 Upon a match, the firewall will create a dynamic rule as if
1826 However, this option doesn't imply an implicit
1830 .It Cm recv | xmit | via Brq Ar ifX | Ar if Ns Cm * | Ar table Ns Po Ar name Ns Oo , Ns Ar value Oc Pc | Ar ipno | Ar any
1831 Matches packets received, transmitted or going through,
1832 respectively, the interface specified by exact name
1836 by IP address, or through some interface.
1839 may be used to match interface by its kernel ifindex.
1842 section below for more information on lookup tables.
1846 keyword causes the interface to always be checked.
1853 then only the receive or transmit interface (respectively)
1855 By specifying both, it is possible to match packets based on
1856 both receive and transmit interface, e.g.:
1858 .Dl "ipfw add deny ip from any to any out recv ed0 xmit ed1"
1862 interface can be tested on either incoming or outgoing packets,
1865 interface can only be tested on outgoing packets.
1870 is invalid) whenever
1874 A packet might not have a receive or transmit interface: packets
1875 originating from the local host have no receive interface,
1876 while packets destined for the local host have no transmit
1878 .It Cm set-limit Bro Cm src-addr | src-port | dst-addr | dst-port Brc Ar N
1881 but does not have an implicit
1885 Matches TCP packets that have the SYN bit set but no ACK bit.
1886 This is the short form of
1887 .Dq Li tcpflags\ syn,!ack .
1889 Matches packets that are associated to a local socket and
1890 for which the SO_USER_COOKIE socket option has been set
1891 to a non-zero value.
1892 As a side effect, the value of the
1893 option is made available as
1895 value, which in turn can be used as
1900 .It Cm src-ip Ar ip-address
1901 Matches IPv4 packets whose source IP is one of the address(es)
1902 specified as an argument.
1903 .It Cm src-ip6 Ar ip6-address
1904 Matches IPv6 packets whose source IP is one of the address(es)
1905 specified as an argument.
1906 .It Cm src-port Ar ports
1907 Matches IP packets whose source port is one of the port(s)
1908 specified as argument.
1909 .It Cm tagged Ar tag-list
1910 Matches packets whose tags are included in
1912 which is either a single value or a list of values or ranges
1913 specified in the same way as
1915 Tags can be applied to the packet using
1917 rule action parameter (see it's description for details on tags).
1918 .It Cm tcpack Ar ack
1920 Match if the TCP header acknowledgment number field is set to
1922 .It Cm tcpdatalen Ar tcpdatalen-list
1923 Matches TCP packets whose length of TCP data is
1924 .Ar tcpdatalen-list ,
1925 which is either a single value or a list of values or ranges
1926 specified in the same way as
1928 .It Cm tcpflags Ar spec
1930 Match if the TCP header contains the comma separated list of
1933 The supported TCP flags are:
1942 The absence of a particular flag may be denoted
1945 A rule which contains a
1947 specification can never match a fragmented packet which has
1951 option for details on matching fragmented packets.
1952 .It Cm tcpseq Ar seq
1954 Match if the TCP header sequence number field is set to
1956 .It Cm tcpwin Ar tcpwin-list
1957 Matches TCP packets whose header window field is set to
1959 which is either a single value or a list of values or ranges
1960 specified in the same way as
1962 .It Cm tcpoptions Ar spec
1964 Match if the TCP header contains the comma separated list of
1965 options specified in
1967 The supported TCP options are:
1970 (maximum segment size),
1972 (tcp window advertisement),
1976 (rfc1323 timestamp) and
1978 (rfc1644 t/tcp connection count).
1979 The absence of a particular option may be denoted
1983 Match all TCP or UDP packets sent by or received for a
1987 may be matched by name or identification number.
1989 For incoming packets,
1990 a routing table lookup is done on the packet's source address.
1991 If the interface on which the packet entered the system matches the
1992 outgoing interface for the route,
1994 If the interfaces do not match up,
1995 the packet does not match.
1996 All outgoing packets or packets with no incoming interface match.
1998 The name and functionality of the option is intentionally similar to
1999 the Cisco IOS command:
2001 .Dl ip verify unicast reverse-path
2003 This option can be used to make anti-spoofing rules to reject all
2004 packets with source addresses not from this interface.
2008 For incoming packets,
2009 a routing table lookup is done on the packet's source address.
2010 If a route to the source address exists, but not the default route
2011 or a blackhole/reject route, the packet matches.
2012 Otherwise, the packet does not match.
2013 All outgoing packets match.
2015 The name and functionality of the option is intentionally similar to
2016 the Cisco IOS command:
2018 .Dl ip verify unicast source reachable-via any
2020 This option can be used to make anti-spoofing rules to reject all
2021 packets whose source address is unreachable.
2023 For incoming packets, the packet's source address is checked if it
2024 belongs to a directly connected network.
2025 If the network is directly connected, then the interface the packet
2026 came on in is compared to the interface the network is connected to.
2027 When incoming interface and directly connected interface are not the
2028 same, the packet does not match.
2029 Otherwise, the packet does match.
2030 All outgoing packets match.
2032 This option can be used to make anti-spoofing rules to reject all
2033 packets that pretend to be from a directly connected network but do
2034 not come in through that interface.
2035 This option is similar to but more restricted than
2037 because it engages only on packets with source addresses of directly
2038 connected networks instead of all source addresses.
2041 Lookup tables are useful to handle large sparse sets of
2042 addresses or other search keys (e.g., ports, jail IDs, interface names).
2043 In the rest of this section we will use the term ``key''.
2044 Table name needs to match the following spec:
2046 Tables with the same name can be created in different
2048 However, rule links to the tables in
2051 This behavior can be controlled by
2052 .Va net.inet.ip.fw.tables_sets
2056 section for more information.
2057 There may be up to 65535 different lookup tables.
2059 The following table types are supported:
2060 .Bl -tag -width indent
2061 .It Ar table-type : Ar addr | iface | number | flow
2062 .It Ar table-key : Ar addr Ns Oo / Ns Ar masklen Oc | iface-name | number | flow-spec
2063 .It Ar flow-spec : Ar flow-field Ns Op , Ns Ar flow-spec
2064 .It Ar flow-field : src-ip | proto | src-port | dst-ip | dst-port
2066 matches IPv4 or IPv6 address.
2067 Each entry is represented by an
2068 .Ar addr Ns Op / Ns Ar masklen
2069 and will match all addresses with base
2071 (specified as an IPv4/IPv6 address, or a hostname) and mask width of
2076 is not specified, it defaults to 32 for IPv4 and 128 for IPv6.
2077 When looking up an IP address in a table, the most specific
2080 matches interface names.
2081 Each entry is represented by string treated as interface name.
2082 Wildcards are not supported.
2084 maches protocol ports, uids/gids or jail IDs.
2085 Each entry is represented by 32-bit unsigned integer.
2086 Ranges are not supported.
2088 Matches packet fields specified by
2090 type suboptions with table entries.
2093 Tables require explicit creation via
2097 The following creation options are supported:
2098 .Bl -tag -width indent
2099 .It Ar create-options : Ar create-option | create-options
2100 .It Ar create-option : Cm type Ar table-type | Cm valtype Ar value-mask | Cm algo Ar algo-desc |
2101 .Cm limit Ar number | Cm locked
2107 Table algorithm to use (see below).
2109 Maximum number of items that may be inserted into table.
2111 Restrict any table modifications.
2114 Some of these options may be modified later via
2117 The following options can be changed:
2118 .Bl -tag -width indent
2119 .It Ar modify-options : Ar modify-option | modify-options
2120 .It Ar modify-option : Cm limit Ar number
2122 Alter maximum number of items that may be inserted into table.
2125 Additionally, table can be locked or unlocked using
2133 can be swapped with each other using
2136 Swap may fail if tables limits are set and data exchange
2137 would result in limits hit.
2138 Operation is performed atomically.
2140 One or more entries can be added to a table at once using
2143 Addition of all items are performed atomically.
2144 By default, error in addition of one entry does not influence
2145 addition of other entries. However, non-zero error code is returned
2149 keyword may be specified before
2151 to indicate all-or-none add request.
2153 One or more entries can be removed from a table at once using
2156 By default, error in removal of one entry does not influence
2157 removing of other entries. However, non-zero error code is returned
2160 It may be possible to check what entry will be found on particular
2166 This functionality is optional and may be unsupported in some algorithms.
2168 The following operations can be performed on
2173 .Bl -tag -width indent
2177 Removes all entries.
2179 Shows generic table information.
2181 Shows generic table information and algo-specific data.
2184 The following lookup algorithms are supported:
2185 .Bl -tag -width indent
2186 .It Ar algo-desc : algo-name | "algo-name algo-data"
2187 .It Ar algo-name: Ar addr:radix | addr:hash | iface:array | number:array | flow:hash
2189 Separate Radix trees for IPv4 and IPv6, the same way as the routing table (see
2195 Separate auto-growing hashes for IPv4 and IPv6.
2196 Accepts entries with the same mask length specified initially via
2197 .Cm "addr:hash masks=/v4,/v6"
2198 algorithm creation options.
2199 Assume /32 and /128 masks by default.
2200 Search removes host bits (according to mask) from supplied address and checks
2201 resulting key in appropriate hash.
2202 Mostly optimized for /64 and byte-ranged IPv6 masks.
2204 Array storing sorted indexes for entries which are presented in the system.
2205 Optimized for very fast lookup.
2207 Array storing sorted u32 numbers.
2209 Auto-growing hash storing flow entries.
2210 Search calculates hash on required packet fields and searches for matching
2211 entries in selected bucket.
2216 feature provides the ability to use a value, looked up in the table, as
2217 the argument for a rule action, action parameter or rule option.
2218 This can significantly reduce number of rules in some configurations.
2219 If two tables are used in a rule, the result of the second (destination)
2222 Each record may hold one or more values according to
2224 This mask is set on table creation via
2227 The following value types are supported:
2228 .Bl -tag -width indent
2229 .It Ar value-mask : Ar value-type Ns Op , Ns Ar value-mask
2230 .It Ar value-type : Ar skipto | pipe | fib | nat | dscp | tag | divert |
2231 .Ar netgraph | limit | ipv4
2233 rule number to jump to.
2237 fib number to match/set.
2239 nat number to jump to.
2241 dscp value to match/set.
2243 tag number to match/set.
2245 port number to divert traffic to.
2247 hook number to move packet to.
2249 maximum number of connections.
2251 IPv4 nexthop to fwd packets to.
2253 IPv6 nexthop to fwd packets to.
2258 argument can be used with the following actions:
2259 .Cm nat, pipe , queue, divert, tee, netgraph, ngtee, fwd, skipto, setfib,
2267 action, the user should be aware that the code will walk the ruleset
2268 up to a rule equal to, or past, the given number.
2272 Section for example usage of tables and the tablearg keyword.
2274 Each rule or table belongs to one of 32 different
2277 Set 31 is reserved for the default rule.
2279 By default, rules or tables are put in set 0, unless you use the
2281 attribute when adding a new rule or table.
2282 Sets can be individually and atomically enabled or disabled,
2283 so this mechanism permits an easy way to store multiple configurations
2284 of the firewall and quickly (and atomically) switch between them.
2286 By default, tables from set 0 are referenced when adding rule with
2287 table opcodes regardless of rule set.
2288 This behavior can be changed by setting
2289 .Va net.inet.ip.fw.tables_sets
2291 Rule's set will then be used for table references.
2293 The command to enable/disable sets is
2294 .Bd -ragged -offset indent
2296 .Cm set Oo Cm disable Ar number ... Oc Op Cm enable Ar number ...
2303 sections can be specified.
2304 Command execution is atomic on all the sets specified in the command.
2305 By default, all sets are enabled.
2307 When you disable a set, its rules behave as if they do not exist
2308 in the firewall configuration, with only one exception:
2309 .Bd -ragged -offset indent
2310 dynamic rules created from a rule before it had been disabled
2311 will still be active until they expire.
2313 dynamic rules you have to explicitly delete the parent rule
2314 which generated them.
2317 The set number of rules can be changed with the command
2318 .Bd -ragged -offset indent
2321 .Brq Cm rule Ar rule-number | old-set
2325 Also, you can atomically swap two rulesets with the command
2326 .Bd -ragged -offset indent
2328 .Cm set swap Ar first-set second-set
2333 Section on some possible uses of sets of rules.
2334 .Sh STATEFUL FIREWALL
2335 Stateful operation is a way for the firewall to dynamically
2336 create rules for specific flows when packets that
2337 match a given pattern are detected.
2338 Support for stateful
2339 operation comes through the
2340 .Cm check-state , keep-state , record-state , limit
2346 Dynamic rules are created when a packet matches a
2352 rule, causing the creation of a
2354 rule which will match all and only packets with
2358 .Em src-ip/src-port dst-ip/dst-port
2363 are used here only to denote the initial match addresses, but they
2364 are completely equivalent afterwards).
2370 This name is used in matching together with addresses, ports and protocol.
2371 Dynamic rules will be checked at the first
2372 .Cm check-state, keep-state
2375 occurrence, and the action performed upon a match will be the same
2376 as in the parent rule.
2378 Note that no additional attributes other than protocol and IP addresses
2379 and ports and :flowname are checked on dynamic rules.
2381 The typical use of dynamic rules is to keep a closed firewall configuration,
2382 but let the first TCP SYN packet from the inside network install a
2383 dynamic rule for the flow so that packets belonging to that session
2384 will be allowed through the firewall:
2386 .Dl "ipfw add check-state :OUTBOUND"
2387 .Dl "ipfw add allow tcp from my-subnet to any setup keep-state :OUTBOUND"
2388 .Dl "ipfw add deny tcp from any to any"
2390 A similar approach can be used for UDP, where an UDP packet coming
2391 from the inside will install a dynamic rule to let the response through
2394 .Dl "ipfw add check-state :OUTBOUND"
2395 .Dl "ipfw add allow udp from my-subnet to any keep-state :OUTBOUND"
2396 .Dl "ipfw add deny udp from any to any"
2398 Dynamic rules expire after some time, which depends on the status
2399 of the flow and the setting of some
2403 .Sx SYSCTL VARIABLES
2405 For TCP sessions, dynamic rules can be instructed to periodically
2406 send keepalive packets to refresh the state of the rule when it is
2411 for more examples on how to use dynamic rules.
2412 .Sh TRAFFIC SHAPER (DUMMYNET) CONFIGURATION
2414 is also the user interface for the
2416 traffic shaper, packet scheduler and network emulator, a subsystem that
2417 can artificially queue, delay or drop packets
2418 emulating the behaviour of certain network links
2419 or queueing systems.
2422 operates by first using the firewall to select packets
2423 using any match pattern that can be used in
2426 Matching packets are then passed to either of two
2427 different objects, which implement the traffic regulation:
2428 .Bl -hang -offset XXXX
2434 with given bandwidth and propagation delay,
2435 driven by a FIFO scheduler and a single queue with programmable
2436 queue size and packet loss rate.
2437 Packets are appended to the queue as they come out from
2439 and then transferred in FIFO order to the link at the desired rate.
2443 is an abstraction used to implement packet scheduling
2444 using one of several packet scheduling algorithms.
2447 are first grouped into flows according to a mask on the 5-tuple.
2448 Flows are then passed to the scheduler associated to the
2450 and each flow uses scheduling parameters (weight and others)
2451 as configured in the
2454 A scheduler in turn is connected to an emulated link,
2455 and arbitrates the link's bandwidth among backlogged flows according to
2456 weights and to the features of the scheduling algorithm in use.
2461 can be used to set hard limits to the bandwidth that a flow can use, whereas
2463 can be used to determine how different flows share the available bandwidth.
2465 A graphical representation of the binding of queues,
2466 flows, schedulers and links is below.
2467 .Bd -literal -offset indent
2468 (flow_mask|sched_mask) sched_mask
2469 +---------+ weight Wx +-------------+
2470 | |->-[flow]-->--| |-+
2471 -->--| QUEUE x | ... | | |
2472 | |->-[flow]-->--| SCHEDuler N | |
2474 ... | +--[LINK N]-->--
2475 +---------+ weight Wy | | +--[LINK N]-->--
2476 | |->-[flow]-->--| | |
2477 -->--| QUEUE y | ... | | |
2478 | |->-[flow]-->--| | |
2479 +---------+ +-------------+ |
2482 It is important to understand the role of the SCHED_MASK
2483 and FLOW_MASK, which are configured through the commands
2484 .Dl "ipfw sched N config mask SCHED_MASK ..."
2486 .Dl "ipfw queue X config mask FLOW_MASK ..." .
2488 The SCHED_MASK is used to assign flows to one or more
2489 scheduler instances, one for each
2490 value of the packet's 5-tuple after applying SCHED_MASK.
2491 As an example, using ``src-ip 0xffffff00'' creates one instance
2492 for each /24 destination subnet.
2494 The FLOW_MASK, together with the SCHED_MASK, is used to split
2496 As an example, using
2497 ``src-ip 0x000000ff''
2498 together with the previous SCHED_MASK makes a flow for
2499 each individual source address.
2500 In turn, flows for each /24
2501 subnet will be sent to the same scheduler instance.
2503 The above diagram holds even for the
2505 case, with the only restriction that a
2507 only supports a SCHED_MASK, and forces the use of a FIFO
2508 scheduler (these are for backward compatibility reasons;
2509 in fact, internally, a
2511 pipe is implemented exactly as above).
2513 There are two modes of
2521 mode tries to emulate a real link: the
2523 scheduler ensures that the packet will not leave the pipe faster than it
2524 would on the real link with a given bandwidth.
2527 mode allows certain packets to bypass the
2529 scheduler (if packet flow does not exceed pipe's bandwidth).
2530 This is the reason why the
2532 mode requires less CPU cycles per packet (on average) and packet latency
2533 can be significantly lower in comparison to a real link with the same
2539 mode can be enabled by setting the
2540 .Va net.inet.ip.dummynet.io_fast
2542 variable to a non-zero value.
2544 .Ss PIPE, QUEUE AND SCHEDULER CONFIGURATION
2550 configuration commands are the following:
2551 .Bd -ragged -offset indent
2552 .Cm pipe Ar number Cm config Ar pipe-configuration
2554 .Cm queue Ar number Cm config Ar queue-configuration
2556 .Cm sched Ar number Cm config Ar sched-configuration
2559 The following parameters can be configured for a pipe:
2561 .Bl -tag -width indent -compact
2562 .It Cm bw Ar bandwidth | device
2563 Bandwidth, measured in
2566 .Brq Cm bit/s | Byte/s .
2569 A value of 0 (default) means unlimited bandwidth.
2570 The unit must immediately follow the number, as in
2572 .Dl "ipfw pipe 1 config bw 300Kbit/s"
2574 If a device name is specified instead of a numeric value, as in
2576 .Dl "ipfw pipe 1 config bw tun0"
2578 then the transmit clock is supplied by the specified device.
2579 At the moment only the
2581 device supports this
2582 functionality, for use in conjunction with
2585 .It Cm delay Ar ms-delay
2586 Propagation delay, measured in milliseconds.
2587 The value is rounded to the next multiple of the clock tick
2588 (typically 10ms, but it is a good practice to run kernels
2590 .Dq "options HZ=1000"
2592 the granularity to 1ms or less).
2593 The default value is 0, meaning no delay.
2595 .It Cm burst Ar size
2596 If the data to be sent exceeds the pipe's bandwidth limit
2597 (and the pipe was previously idle), up to
2599 bytes of data are allowed to bypass the
2601 scheduler, and will be sent as fast as the physical link allows.
2602 Any additional data will be transmitted at the rate specified
2606 The burst size depends on how long the pipe has been idle;
2607 the effective burst size is calculated as follows:
2614 .It Cm profile Ar filename
2615 A file specifying the additional overhead incurred in the transmission
2616 of a packet on the link.
2618 Some link types introduce extra delays in the transmission
2619 of a packet, e.g., because of MAC level framing, contention on
2620 the use of the channel, MAC level retransmissions and so on.
2621 From our point of view, the channel is effectively unavailable
2622 for this extra time, which is constant or variable depending
2624 Additionally, packets may be dropped after this
2625 time (e.g., on a wireless link after too many retransmissions).
2626 We can model the additional delay with an empirical curve
2627 that represents its distribution.
2628 .Bd -literal -offset indent
2629 cumulative probability
2639 +-------*------------------->
2642 The empirical curve may have both vertical and horizontal lines.
2643 Vertical lines represent constant delay for a range of
2645 Horizontal lines correspond to a discontinuity in the delay
2646 distribution: the pipe will use the largest delay for a
2649 The file format is the following, with whitespace acting as
2650 a separator and '#' indicating the beginning a comment:
2651 .Bl -tag -width indent
2652 .It Cm name Ar identifier
2653 optional name (listed by "ipfw pipe show")
2654 to identify the delay distribution;
2656 the bandwidth used for the pipe.
2657 If not specified here, it must be present
2658 explicitly as a configuration parameter for the pipe;
2659 .It Cm loss-level Ar L
2660 the probability above which packets are lost.
2661 (0.0 <= L <= 1.0, default 1.0 i.e., no loss);
2663 the number of samples used in the internal
2664 representation of the curve (2..1024; default 100);
2665 .It Cm "delay prob" | "prob delay"
2666 One of these two lines is mandatory and defines
2667 the format of the following lines with data points.
2669 2 or more lines representing points in the curve,
2670 with either delay or probability first, according
2671 to the chosen format.
2672 The unit for delay is milliseconds.
2673 Data points do not need to be sorted.
2674 Also, the number of actual lines can be different
2675 from the value of the "samples" parameter:
2677 utility will sort and interpolate
2678 the curve as needed.
2681 Example of a profile file:
2682 .Bd -literal -offset indent
2687 0 200 # minimum overhead is 200ms
2693 #configuration file end
2697 The following parameters can be configured for a queue:
2699 .Bl -tag -width indent -compact
2700 .It Cm pipe Ar pipe_nr
2701 Connects a queue to the specified pipe.
2702 Multiple queues (with the same or different weights) can be connected to
2703 the same pipe, which specifies the aggregate rate for the set of queues.
2705 .It Cm weight Ar weight
2706 Specifies the weight to be used for flows matching this queue.
2707 The weight must be in the range 1..100, and defaults to 1.
2710 The following case-insensitive parameters can be configured for a
2713 .Bl -tag -width indent -compact
2714 .It Cm type Ar {fifo | wf2q+ | rr | qfq}
2715 specifies the scheduling algorithm to use.
2716 .Bl -tag -width indent -compact
2718 is just a FIFO scheduler (which means that all packets
2719 are stored in the same queue as they arrive to the scheduler).
2720 FIFO has O(1) per-packet time complexity, with very low
2721 constants (estimate 60-80ns on a 2GHz desktop machine)
2722 but gives no service guarantees.
2724 implements the WF2Q+ algorithm, which is a Weighted Fair Queueing
2725 algorithm which permits flows to share bandwidth according to
2727 Note that weights are not priorities; even a flow
2728 with a minuscule weight will never starve.
2729 WF2Q+ has O(log N) per-packet processing cost, where N is the number
2730 of flows, and is the default algorithm used by previous versions
2733 implements the Deficit Round Robin algorithm, which has O(1) processing
2734 costs (roughly, 100-150ns per packet)
2735 and permits bandwidth allocation according to weights, but
2736 with poor service guarantees.
2738 implements the QFQ algorithm, which is a very fast variant of
2739 WF2Q+, with similar service guarantees and O(1) processing
2740 costs (roughly, 200-250ns per packet).
2744 In addition to the type, all parameters allowed for a pipe can also
2745 be specified for a scheduler.
2747 Finally, the following parameters can be configured for both
2750 .Bl -tag -width XXXX -compact
2751 .It Cm buckets Ar hash-table-size
2752 Specifies the size of the hash table used for storing the
2754 Default value is 64 controlled by the
2757 .Va net.inet.ip.dummynet.hash_size ,
2758 allowed range is 16 to 65536.
2760 .It Cm mask Ar mask-specifier
2761 Packets sent to a given pipe or queue by an
2763 rule can be further classified into multiple flows, each of which is then
2767 A flow identifier is constructed by masking the IP addresses,
2768 ports and protocol types as specified with the
2770 options in the configuration of the pipe or queue.
2771 For each different flow identifier, a new pipe or queue is created
2772 with the same parameters as the original object, and matching packets
2777 are used, each flow will get the same bandwidth as defined by the pipe,
2780 are used, each flow will share the parent's pipe bandwidth evenly
2781 with other flows generated by the same queue (note that other queues
2782 with different weights might be connected to the same pipe).
2784 Available mask specifiers are a combination of one or more of the following:
2786 .Cm dst-ip Ar mask ,
2787 .Cm dst-ip6 Ar mask ,
2788 .Cm src-ip Ar mask ,
2789 .Cm src-ip6 Ar mask ,
2790 .Cm dst-port Ar mask ,
2791 .Cm src-port Ar mask ,
2792 .Cm flow-id Ar mask ,
2797 where the latter means all bits in all fields are significant.
2800 When a packet is dropped by a
2802 queue or pipe, the error
2803 is normally reported to the caller routine in the kernel, in the
2804 same way as it happens when a device queue fills up.
2806 option reports the packet as successfully delivered, which can be
2807 needed for some experimental setups where you want to simulate
2808 loss or congestion at a remote router.
2810 .It Cm plr Ar packet-loss-rate
2813 .Ar packet-loss-rate
2814 is a floating-point number between 0 and 1, with 0 meaning no
2815 loss, 1 meaning 100% loss.
2816 The loss rate is internally represented on 31 bits.
2818 .It Cm queue Brq Ar slots | size Ns Cm Kbytes
2823 Default value is 50 slots, which
2824 is the typical queue size for Ethernet devices.
2825 Note that for slow speed links you should keep the queue
2826 size short or your traffic might be affected by a significant
2828 E.g., 50 max-sized ethernet packets (1500 bytes) mean 600Kbit
2829 or 20s of queue on a 30Kbit/s pipe.
2830 Even worse effects can result if you get packets from an
2831 interface with a much larger MTU, e.g.\& the loopback interface
2832 with its 16KB packets.
2836 .Em net.inet.ip.dummynet.pipe_byte_limit
2838 .Em net.inet.ip.dummynet.pipe_slot_limit
2839 control the maximum lengths that can be specified.
2841 .It Cm red | gred Ar w_q Ns / Ns Ar min_th Ns / Ns Ar max_th Ns / Ns Ar max_p
2843 Make use of the RED (Random Early Detection) queue management algorithm.
2848 point numbers between 0 and 1 (inclusive), while
2852 are integer numbers specifying thresholds for queue management
2853 (thresholds are computed in bytes if the queue has been defined
2854 in bytes, in slots otherwise).
2855 The two parameters can also be of the same value if needed. The
2857 also supports the gentle RED variant (gred) and ECN (Explicit Congestion
2858 Notification) as optional. Three
2860 variables can be used to control the RED behaviour:
2861 .Bl -tag -width indent
2862 .It Va net.inet.ip.dummynet.red_lookup_depth
2863 specifies the accuracy in computing the average queue
2864 when the link is idle (defaults to 256, must be greater than zero)
2865 .It Va net.inet.ip.dummynet.red_avg_pkt_size
2866 specifies the expected average packet size (defaults to 512, must be
2868 .It Va net.inet.ip.dummynet.red_max_pkt_size
2869 specifies the expected maximum packet size, only used when queue
2870 thresholds are in bytes (defaults to 1500, must be greater than zero).
2874 When used with IPv6 data,
2876 currently has several limitations.
2877 Information necessary to route link-local packets to an
2878 interface is not available after processing by
2880 so those packets are dropped in the output path.
2881 Care should be taken to ensure that link-local packets are not passed to
2884 Here are some important points to consider when designing your
2888 Remember that you filter both packets going
2892 Most connections need packets going in both directions.
2894 Remember to test very carefully.
2895 It is a good idea to be near the console when doing this.
2896 If you cannot be near the console,
2897 use an auto-recovery script such as the one in
2898 .Pa /usr/share/examples/ipfw/change_rules.sh .
2900 Do not forget the loopback interface.
2905 There are circumstances where fragmented datagrams are unconditionally
2907 TCP packets are dropped if they do not contain at least 20 bytes of
2908 TCP header, UDP packets are dropped if they do not contain a full 8
2909 byte UDP header, and ICMP packets are dropped if they do not contain
2910 4 bytes of ICMP header, enough to specify the ICMP type, code, and
2912 These packets are simply logged as
2914 since there may not be enough good data in the packet to produce a
2915 meaningful log entry.
2917 Another type of packet is unconditionally dropped, a TCP packet with a
2918 fragment offset of one.
2919 This is a valid packet, but it only has one use, to try
2920 to circumvent firewalls.
2921 When logging is enabled, these packets are
2922 reported as being dropped by rule -1.
2924 If you are logged in over a network, loading the
2928 is probably not as straightforward as you would think.
2929 The following command line is recommended:
2930 .Bd -literal -offset indent
2932 ipfw add 32000 allow ip from any to any
2935 Along the same lines, doing an
2936 .Bd -literal -offset indent
2940 in similar surroundings is also a bad idea.
2944 filter list may not be modified if the system security level
2945 is set to 3 or higher
2948 for information on system security levels).
2950 .Sh PACKET DIVERSION
2953 socket bound to the specified port will receive all packets
2954 diverted to that port.
2955 If no socket is bound to the destination port, or if the divert module is
2956 not loaded, or if the kernel was not compiled with divert socket support,
2957 the packets are dropped.
2958 .Sh NETWORK ADDRESS TRANSLATION (NAT)
2960 support in-kernel NAT using the kernel version of
2964 should be loaded or kernel should have
2965 .Cm options IPFIREWALL_NAT
2968 The nat configuration command is the following:
2969 .Bd -ragged -offset indent
2974 .Ar nat-configuration
2978 The following parameters can be configured:
2979 .Bl -tag -width indent
2980 .It Cm ip Ar ip_address
2981 Define an ip address to use for aliasing.
2983 Use ip address of NIC for aliasing, dynamically changing
2984 it if NIC's ip address changes.
2986 Enable logging on this nat instance.
2988 Deny any incoming connection from outside world.
2990 Try to leave the alias port numbers unchanged from
2991 the actual local port numbers.
2993 Traffic on the local network not originating from an
2994 unregistered address spaces will be ignored.
2996 Reset table of the packet aliasing engine on address change.
2998 Reverse the way libalias handles aliasing.
3000 Obey transparent proxy rules only, packet aliasing is not performed.
3002 Skip instance in case of global state lookup (see below).
3005 Some specials value can be supplied instead of
3007 .Bl -tag -width indent
3009 Looks up translation state in all configured nat instances.
3010 If an entry is found, packet is aliased according to that entry.
3011 If no entry was found in any of the instances, packet is passed unchanged,
3012 and no new entry will be created.
3014 .Sx MULTIPLE INSTANCES
3017 for more information.
3019 Uses argument supplied in lookup table.
3022 section below for more information on lookup tables.
3025 To let the packet continue after being (de)aliased, set the sysctl variable
3026 .Va net.inet.ip.fw.one_pass
3028 For more information about aliasing modes, refer to
3032 for some examples about nat usage.
3033 .Ss REDIRECT AND LSNAT SUPPORT IN IPFW
3034 Redirect and LSNAT support follow closely the syntax used in
3038 for some examples on how to do redirect and lsnat.
3039 .Ss SCTP NAT SUPPORT
3040 SCTP nat can be configured in a similar manner to TCP through the
3043 The main difference is that
3045 does not do port translation.
3046 Since the local and global side ports will be the same,
3047 there is no need to specify both.
3048 Ports are redirected as follows:
3049 .Bd -ragged -offset indent
3055 .Cm redirect_port sctp
3056 .Ar ip_address [,addr_list] {[port | port-port] [,ports]}
3062 configuration can be done in real-time through the
3065 All may be changed dynamically, though the hash_table size will only
3070 .Sx SYSCTL VARIABLES
3072 .Sh IPv6/IPv4 NETWORK ADDRESS AND PROTOCOL TRANSLATION
3074 supports in-kernel IPv6/IPv4 network address and protocol translation.
3075 Stateful NAT64 translation allows IPv6-only clients to contact IPv4 servers
3076 using unicast TCP, UDP or ICMP protocols.
3077 One or more IPv4 addresses assigned to a stateful NAT64 translator are shared
3078 among several IPv6-only clients.
3079 When stateful NAT64 is used in conjunction with DNS64, no changes are usually
3080 required in the IPv6 client or the IPv4 server.
3083 should be loaded or kernel should have
3084 .Cm options IPFIREWALL_NAT64
3085 to be able use stateful NAT64 translator.
3087 Stateful NAT64 uses a bunch of memory for several types of objects.
3088 When IPv6 client initiates connection, NAT64 translator creates a host entry
3089 in the states table.
3090 Each host entry has a number of ports group entries allocated on demand.
3091 Ports group entries contains connection state entries.
3092 There are several options to control limits and lifetime for these objects.
3094 NAT64 translator follows RFC7915 when does ICMPv6/ICMP translation,
3095 unsupported message types will be silently dropped.
3096 IPv6 needs several ICMPv6 message types to be explicitly allowed for correct
3098 Make sure that ND6 neighbor solicitation (ICMPv6 type 135) and neighbor
3099 advertisement (ICMPv6 type 136) messages will not be handled by translation
3102 After translation NAT64 translator sends packets through corresponding netisr
3104 Thus translator host should be configured as IPv4 and IPv6 router.
3106 The stateful NAT64 configuration command is the following:
3107 .Bd -ragged -offset indent
3116 The following parameters can be configured:
3117 .Bl -tag -width indent
3118 .It Cm prefix4 Ar ipv4_prefix/plen
3119 The IPv4 prefix with mask defines the pool of IPv4 addresses used as
3120 source address after translation.
3121 Stateful NAT64 module translates IPv6 source address of client to one
3122 IPv4 address from this pool.
3123 Note that incoming IPv4 packets that don't have corresponding state entry
3124 in the states table will be dropped by translator.
3125 Make sure that translation rules handle packets, destined to configured prefix.
3126 .It Cm prefix6 Ar ipv6_prefix/length
3127 The IPv6 prefix defines IPv4-embedded IPv6 addresses used by translator
3128 to represent IPv4 addresses. This IPv6 prefix should be configured in DNS64.
3129 The translator implementation follows RFC6052, that restricts the length of
3130 prefixes to one of following: 32, 40, 48, 56, 64, or 96.
3131 The Well-Known IPv6 Prefix 64:ff9b:: must be 96 bits long.
3132 .It Cm max_ports Ar number
3133 Maximum number of ports reserved for upper level protocols to one IPv6 client.
3134 All reserved ports are divided into chunks between supported protocols.
3135 The number of connections from one IPv6 client is limited by this option.
3136 Note that closed TCP connections still remain in the list of connections until
3138 interval will not expire.
3141 .It Cm host_del_age Ar seconds
3142 The number of seconds until the host entry for a IPv6 client will be deleted
3143 and all its resources will be released due to inactivity.
3146 .It Cm pg_del_age Ar seconds
3147 The number of seconds until a ports group with unused state entries will
3151 .It Cm tcp_syn_age Ar seconds
3152 The number of seconds while a state entry for TCP connection with only SYN
3154 If TCP connection establishing will not be finished,
3155 state entry will be deleted.
3158 .It Cm tcp_est_age Ar seconds
3159 The number of seconds while a state entry for established TCP connection
3163 .It Cm tcp_close_age Ar seconds
3164 The number of seconds while a state entry for closed TCP connection
3166 Keeping state entries for closed connections is needed, because IPv4 servers
3167 typically keep closed connections in a TIME_WAIT state for a several minutes.
3168 Since translator's IPv4 addresses are shared among all IPv6 clients,
3169 new connections from the same addresses and ports may be rejected by server,
3170 because these connections are still in a TIME_WAIT state.
3171 Keeping them in translator's state table protects from such rejects.
3174 .It Cm udp_age Ar seconds
3175 The number of seconds while translator keeps state entry in a waiting for
3176 reply to the sent UDP datagram.
3179 .It Cm icmp_age Ar seconds
3180 The number of seconds while translator keeps state entry in a waiting for
3181 reply to the sent ICMP message.
3185 Turn on logging of all handled packets via BPF through
3189 is a pseudo interface and can be created after a boot manually with
3192 Note that it has different purpose than
3195 Translators sends to BPF an additional information with each packet.
3198 you are able to see each handled packet before and after translation.
3200 Turn off logging of all handled packets via BPF.
3203 To inspect a states table of stateful NAT64 the following command can be used:
3204 .Bd -ragged -offset indent
3213 Stateless NAT64 translator doesn't use a states table for translation
3214 and converts IPv4 addresses to IPv6 and vice versa solely based on the
3215 mappings taken from configured lookup tables.
3216 Since a states table doesn't used by stateless translator,
3217 it can be configured to pass IPv4 clients to IPv6-only servers.
3219 The stateless NAT64 configuration command is the following:
3220 .Bd -ragged -offset indent
3229 The following parameters can be configured:
3230 .Bl -tag -width indent
3231 .It Cm prefix6 Ar ipv6_prefix/length
3232 The IPv6 prefix defines IPv4-embedded IPv6 addresses used by translator
3233 to represent IPv4 addresses. This IPv6 prefix should be configured in DNS64.
3234 .It Cm table4 Ar table46
3237 contains mapping how IPv4 addresses should be translated to IPv6 addresses.
3238 .It Cm table6 Ar table64
3241 contains mapping how IPv6 addresses should be translated to IPv4 addresses.
3243 Turn on logging of all handled packets via BPF through
3247 Turn off logging of all handled packets via BPF.
3250 Note that the behavior of stateless translator with respect to not matched
3251 packets differs from stateful translator.
3252 If corresponding addresses was not found in the lookup tables, the packet
3253 will not be dropped and the search continues.
3254 .Sh IPv6-to-IPv6 NETWORK PREFIX TRANSLATION (NPTv6)
3256 supports in-kernel IPv6-to-IPv6 network prefix translation as described
3260 should be loaded or kernel should has
3261 .Cm options IPFIREWALL_NPTV6
3262 to be able use NPTv6 translator.
3264 The NPTv6 configuration command is the following:
3265 .Bd -ragged -offset indent
3274 The following parameters can be configured:
3275 .Bl -tag -width indent
3276 .It Cm int_prefix Ar ipv6_prefix
3277 IPv6 prefix used in internal network.
3278 NPTv6 module translates source address when it matches this prefix.
3279 .It Cm ext_prefix Ar ipv6_prefix
3280 IPv6 prefix used in external network.
3281 NPTv6 module translates destination address when it matches this prefix.
3282 .It Cm prefixlen Ar length
3283 The length of specified IPv6 prefixes. It must be in range from 8 to 64.
3286 Note that the prefix translation rules are silently ignored when IPv6 packet
3287 forwarding is disabled.
3288 To enable the packet forwarding, set the sysctl variable
3289 .Va net.inet6.ip6.forwarding
3292 To let the packet continue after being translated, set the sysctl variable
3293 .Va net.inet.ip.fw.one_pass
3296 Tunables can be set in
3302 before ipfw module gets loaded.
3303 .Bl -tag -width indent
3304 .It Va net.inet.ip.fw.default_to_accept: No 0
3305 Defines ipfw last rule behavior.
3306 This value overrides
3307 .Cd "options IPFW_DEFAULT_TO_(ACCEPT|DENY)"
3308 from kernel configuration file.
3309 .It Va net.inet.ip.fw.tables_max: No 128
3310 Defines number of tables available in ipfw.
3311 Number cannot exceed 65534.
3313 .Sh SYSCTL VARIABLES
3316 variables controls the behaviour of the firewall and
3318 .Pq Nm dummynet , bridge , sctp nat .
3319 These are shown below together with their default value
3320 (but always check with the
3322 command what value is actually in use) and meaning:
3323 .Bl -tag -width indent
3324 .It Va net.inet.ip.alias.sctp.accept_global_ootb_addip: No 0
3327 responds to receipt of global OOTB ASCONF-AddIP:
3328 .Bl -tag -width indent
3330 No response (unless a partially matching association exists -
3331 ports and vtags match but global address does not)
3334 will accept and process all OOTB global AddIP messages.
3337 Option 1 should never be selected as this forms a security risk.
3339 establish multiple fake associations by sending AddIP messages.
3340 .It Va net.inet.ip.alias.sctp.chunk_proc_limit: No 5
3341 Defines the maximum number of chunks in an SCTP packet that will be
3343 packet that matches an existing association.
3344 This value is enforced to be greater or equal than
3345 .Cm net.inet.ip.alias.sctp.initialising_chunk_proc_limit .
3347 a DoS risk yet setting too low a value may result in
3348 important control chunks in
3349 the packet not being located and parsed.
3350 .It Va net.inet.ip.alias.sctp.error_on_ootb: No 1
3353 responds to any Out-of-the-Blue (OOTB) packets with ErrorM packets.
3354 An OOTB packet is a packet that arrives with no existing association
3357 and is not an INIT or ASCONF-AddIP packet:
3358 .Bl -tag -width indent
3360 ErrorM is never sent in response to OOTB packets.
3362 ErrorM is only sent to OOTB packets received on the local side.
3364 ErrorM is sent to the local side and on the global side ONLY if there is a
3365 partial match (ports and vtags match but the source global IP does not).
3366 This value is only useful if the
3368 is tracking global IP addresses.
3370 ErrorM is sent in response to all OOTB packets on both
3371 the local and global side
3375 At the moment the default is 0, since the ErrorM packet is not yet
3376 supported by most SCTP stacks.
3377 When it is supported, and if not tracking
3378 global addresses, we recommend setting this value to 1 to allow
3379 multi-homed local hosts to function with the
3381 To track global addresses, we recommend setting this value to 2 to
3382 allow global hosts to be informed when they need to (re)send an
3384 Value 3 should never be chosen (except for debugging) as the
3386 will respond to all OOTB global packets (a DoS risk).
3387 .It Va net.inet.ip.alias.sctp.hashtable_size: No 2003
3388 Size of hash tables used for
3390 lookups (100 < prime_number > 1000001).
3393 size for any future created
3395 instance and therefore must be set prior to creating a
3398 The table sizes may be changed to suit specific needs.
3399 If there will be few
3400 concurrent associations, and memory is scarce, you may make these smaller.
3401 If there will be many thousands (or millions) of concurrent associations, you
3402 should make these larger.
3403 A prime number is best for the table size.
3405 update function will adjust your input value to the next highest prime number.
3406 .It Va net.inet.ip.alias.sctp.holddown_time: No 0
3407 Hold association in table for this many seconds after receiving a
3409 This allows endpoints to correct shutdown gracefully if a
3410 shutdown_complete is lost and retransmissions are required.
3411 .It Va net.inet.ip.alias.sctp.init_timer: No 15
3412 Timeout value while waiting for (INIT-ACK|AddIP-ACK).
3413 This value cannot be 0.
3414 .It Va net.inet.ip.alias.sctp.initialising_chunk_proc_limit: No 2
3415 Defines the maximum number of chunks in an SCTP packet that will be parsed when
3416 no existing association exists that matches that packet.
3418 will only be an INIT or ASCONF-AddIP packet.
3419 A higher value may become a DoS
3420 risk as malformed packets can consume processing resources.
3421 .It Va net.inet.ip.alias.sctp.param_proc_limit: No 25
3422 Defines the maximum number of parameters within a chunk that will be
3425 As for other similar sysctl variables, larger values pose a DoS risk.
3426 .It Va net.inet.ip.alias.sctp.log_level: No 0
3427 Level of detail in the system log messages (0 \- minimal, 1 \- event,
3428 2 \- info, 3 \- detail, 4 \- debug, 5 \- max debug).
3430 option in high loss environments.
3431 .It Va net.inet.ip.alias.sctp.shutdown_time: No 15
3432 Timeout value while waiting for SHUTDOWN-COMPLETE.
3433 This value cannot be 0.
3434 .It Va net.inet.ip.alias.sctp.track_global_addresses: No 0
3435 Enables/disables global IP address tracking within the
3438 upper limit on the number of addresses tracked for each association:
3439 .Bl -tag -width indent
3441 Global tracking is disabled
3443 Enables tracking, the maximum number of addresses tracked for each
3444 association is limited to this value
3447 This variable is fully dynamic, the new value will be adopted for all newly
3448 arriving associations, existing associations are treated
3449 as they were previously.
3450 Global tracking will decrease the number of collisions within the
3453 of increased processing load, memory usage, complexity, and possible
3456 problems in complex networks with multiple
3458 We recommend not tracking
3459 global IP addresses, this will still result in a fully functional
3461 .It Va net.inet.ip.alias.sctp.up_timer: No 300
3462 Timeout value to keep an association up with no traffic.
3463 This value cannot be 0.
3464 .It Va net.inet.ip.dummynet.expire : No 1
3465 Lazily delete dynamic pipes/queue once they have no pending traffic.
3466 You can disable this by setting the variable to 0, in which case
3467 the pipes/queues will only be deleted when the threshold is reached.
3468 .It Va net.inet.ip.dummynet.hash_size : No 64
3469 Default size of the hash table used for dynamic pipes/queues.
3470 This value is used when no
3472 option is specified when configuring a pipe/queue.
3473 .It Va net.inet.ip.dummynet.io_fast : No 0
3474 If set to a non-zero value,
3479 operation (see above) is enabled.
3480 .It Va net.inet.ip.dummynet.io_pkt
3481 Number of packets passed to
3483 .It Va net.inet.ip.dummynet.io_pkt_drop
3484 Number of packets dropped by
3486 .It Va net.inet.ip.dummynet.io_pkt_fast
3487 Number of packets bypassed by the
3490 .It Va net.inet.ip.dummynet.max_chain_len : No 16
3491 Target value for the maximum number of pipes/queues in a hash bucket.
3493 .Cm max_chain_len*hash_size
3494 is used to determine the threshold over which empty pipes/queues
3495 will be expired even when
3496 .Cm net.inet.ip.dummynet.expire=0 .
3497 .It Va net.inet.ip.dummynet.red_lookup_depth : No 256
3498 .It Va net.inet.ip.dummynet.red_avg_pkt_size : No 512
3499 .It Va net.inet.ip.dummynet.red_max_pkt_size : No 1500
3500 Parameters used in the computations of the drop probability
3501 for the RED algorithm.
3502 .It Va net.inet.ip.dummynet.pipe_byte_limit : No 1048576
3503 .It Va net.inet.ip.dummynet.pipe_slot_limit : No 100
3504 The maximum queue size that can be specified in bytes or packets.
3505 These limits prevent accidental exhaustion of resources such as mbufs.
3506 If you raise these limits,
3507 you should make sure the system is configured so that sufficient resources
3509 .It Va net.inet.ip.fw.autoinc_step : No 100
3510 Delta between rule numbers when auto-generating them.
3511 The value must be in the range 1..1000.
3512 .It Va net.inet.ip.fw.curr_dyn_buckets : Va net.inet.ip.fw.dyn_buckets
3513 The current number of buckets in the hash table for dynamic rules
3515 .It Va net.inet.ip.fw.debug : No 1
3516 Controls debugging messages produced by
3518 .It Va net.inet.ip.fw.default_rule : No 65535
3519 The default rule number (read-only).
3521 .Nm , the default rule is the last one, so its number
3522 can also serve as the highest number allowed for a rule.
3523 .It Va net.inet.ip.fw.dyn_buckets : No 256
3524 The number of buckets in the hash table for dynamic rules.
3525 Must be a power of 2, up to 65536.
3526 It only takes effect when all dynamic rules have expired, so you
3527 are advised to use a
3529 command to make sure that the hash table is resized.
3530 .It Va net.inet.ip.fw.dyn_count : No 3
3531 Current number of dynamic rules
3533 .It Va net.inet.ip.fw.dyn_keepalive : No 1
3534 Enables generation of keepalive packets for
3536 rules on TCP sessions.
3537 A keepalive is generated to both
3538 sides of the connection every 5 seconds for the last 20
3539 seconds of the lifetime of the rule.
3540 .It Va net.inet.ip.fw.dyn_max : No 8192
3541 Maximum number of dynamic rules.
3542 When you hit this limit, no more dynamic rules can be
3543 installed until old ones expire.
3544 .It Va net.inet.ip.fw.dyn_ack_lifetime : No 300
3545 .It Va net.inet.ip.fw.dyn_syn_lifetime : No 20
3546 .It Va net.inet.ip.fw.dyn_fin_lifetime : No 1
3547 .It Va net.inet.ip.fw.dyn_rst_lifetime : No 1
3548 .It Va net.inet.ip.fw.dyn_udp_lifetime : No 5
3549 .It Va net.inet.ip.fw.dyn_short_lifetime : No 30
3550 These variables control the lifetime, in seconds, of dynamic
3552 Upon the initial SYN exchange the lifetime is kept short,
3553 then increased after both SYN have been seen, then decreased
3554 again during the final FIN exchange or when a RST is received.
3556 .Em dyn_fin_lifetime
3558 .Em dyn_rst_lifetime
3559 must be strictly lower than 5 seconds, the period of
3560 repetition of keepalives.
3561 The firewall enforces that.
3562 .It Va net.inet.ip.fw.dyn_keep_states: No 0
3563 Keep dynamic states on rule/set deletion.
3564 States are relinked to default rule (65535).
3565 This can be handly for ruleset reload.
3566 Turned off by default.
3567 .It Va net.inet.ip.fw.enable : No 1
3568 Enables the firewall.
3569 Setting this variable to 0 lets you run your machine without
3570 firewall even if compiled in.
3571 .It Va net.inet6.ip6.fw.enable : No 1
3572 provides the same functionality as above for the IPv6 case.
3573 .It Va net.inet.ip.fw.one_pass : No 1
3574 When set, the packet exiting from the
3578 node is not passed though the firewall again.
3579 Otherwise, after an action, the packet is
3580 reinjected into the firewall at the next rule.
3581 .It Va net.inet.ip.fw.tables_max : No 128
3582 Maximum number of tables.
3583 .It Va net.inet.ip.fw.verbose : No 1
3584 Enables verbose messages.
3585 .It Va net.inet.ip.fw.verbose_limit : No 0
3586 Limits the number of messages produced by a verbose firewall.
3587 .It Va net.inet6.ip6.fw.deny_unknown_exthdrs : No 1
3588 If enabled packets with unknown IPv6 Extension Headers will be denied.
3589 .It Va net.link.ether.ipfw : No 0
3590 Controls whether layer-2 packets are passed to
3593 .It Va net.link.bridge.ipfw : No 0
3594 Controls whether bridged packets are passed to
3598 .Sh INTERNAL DIAGNOSTICS
3599 There are some commands that may be useful to understand current state
3600 of certain subsystems inside kernel module.
3601 These commands provide debugging output which may change without notice.
3603 Currently the following commands are available as
3606 .Bl -tag -width indent
3608 Lists all interface which are currently tracked by
3610 with their in-kernel status.
3612 List all table lookup algorithms currently available.
3615 There are far too many possible uses of
3617 so this Section will only give a small set of examples.
3619 .Ss BASIC PACKET FILTERING
3620 This command adds an entry which denies all tcp packets from
3621 .Em cracker.evil.org
3622 to the telnet port of
3624 from being forwarded by the host:
3626 .Dl "ipfw add deny tcp from cracker.evil.org to wolf.tambov.su telnet"
3628 This one disallows any connection from the entire cracker's
3631 .Dl "ipfw add deny ip from 123.45.67.0/24 to my.host.org"
3633 A first and efficient way to limit access (not using dynamic rules)
3634 is the use of the following rules:
3636 .Dl "ipfw add allow tcp from any to any established"
3637 .Dl "ipfw add allow tcp from net1 portlist1 to net2 portlist2 setup"
3638 .Dl "ipfw add allow tcp from net3 portlist3 to net3 portlist3 setup"
3640 .Dl "ipfw add deny tcp from any to any"
3642 The first rule will be a quick match for normal TCP packets,
3643 but it will not match the initial SYN packet, which will be
3646 rules only for selected source/destination pairs.
3647 All other SYN packets will be rejected by the final
3651 If you administer one or more subnets, you can take advantage
3652 of the address sets and or-blocks and write extremely
3653 compact rulesets which selectively enable services to blocks
3654 of clients, as below:
3656 .Dl "goodguys=\*q{ 10.1.2.0/24{20,35,66,18} or 10.2.3.0/28{6,3,11} }\*q"
3657 .Dl "badguys=\*q10.1.2.0/24{8,38,60}\*q"
3659 .Dl "ipfw add allow ip from ${goodguys} to any"
3660 .Dl "ipfw add deny ip from ${badguys} to any"
3661 .Dl "... normal policies ..."
3665 option could be used to do automated anti-spoofing by adding the
3666 following to the top of a ruleset:
3668 .Dl "ipfw add deny ip from any to any not verrevpath in"
3670 This rule drops all incoming packets that appear to be coming to the
3671 system on the wrong interface.
3672 For example, a packet with a source
3673 address belonging to a host on a protected internal network would be
3674 dropped if it tried to enter the system from an external interface.
3678 option could be used to do similar but more restricted anti-spoofing
3679 by adding the following to the top of a ruleset:
3681 .Dl "ipfw add deny ip from any to any not antispoof in"
3683 This rule drops all incoming packets that appear to be coming from another
3684 directly connected system but on the wrong interface.
3685 For example, a packet with a source address of
3686 .Li 192.168.0.0/24 ,
3695 option could be used to (re)mark user traffic,
3696 by adding the following to the appropriate place in ruleset:
3698 .Dl "ipfw add setdscp be ip from any to any dscp af11,af21"
3700 In order to protect a site from flood attacks involving fake
3701 TCP packets, it is safer to use dynamic rules:
3703 .Dl "ipfw add check-state"
3704 .Dl "ipfw add deny tcp from any to any established"
3705 .Dl "ipfw add allow tcp from my-net to any setup keep-state"
3707 This will let the firewall install dynamic rules only for
3708 those connection which start with a regular SYN packet coming
3709 from the inside of our network.
3710 Dynamic rules are checked when encountering the first
3719 rule should usually be placed near the beginning of the
3720 ruleset to minimize the amount of work scanning the ruleset.
3721 Your mileage may vary.
3723 For more complex scenarios with dynamic rules
3727 can be used to precisely control creation and checking of dynamic rules.
3728 Example of usage of these options are provided in
3729 .Sx NETWORK ADDRESS TRANSLATION (NAT)
3732 To limit the number of connections a user can open
3733 you can use the following type of rules:
3735 .Dl "ipfw add allow tcp from my-net/24 to any setup limit src-addr 10"
3736 .Dl "ipfw add allow tcp from any to me setup limit src-addr 4"
3738 The former (assuming it runs on a gateway) will allow each host
3739 on a /24 network to open at most 10 TCP connections.
3740 The latter can be placed on a server to make sure that a single
3741 client does not use more than 4 simultaneous connections.
3744 stateful rules can be subject to denial-of-service attacks
3745 by a SYN-flood which opens a huge number of dynamic rules.
3746 The effects of such attacks can be partially limited by
3749 variables which control the operation of the firewall.
3751 Here is a good usage of the
3753 command to see accounting records and timestamp information:
3757 or in short form without timestamps:
3761 which is equivalent to:
3765 Next rule diverts all incoming packets from 192.168.2.0/24
3766 to divert port 5000:
3768 .Dl ipfw divert 5000 ip from 192.168.2.0/24 to any in
3770 The following rules show some of the applications of
3774 for simulations and the like.
3776 This rule drops random incoming packets with a probability
3779 .Dl "ipfw add prob 0.05 deny ip from any to any in"
3781 A similar effect can be achieved making use of
3785 .Dl "ipfw add pipe 10 ip from any to any"
3786 .Dl "ipfw pipe 10 config plr 0.05"
3788 We can use pipes to artificially limit bandwidth, e.g.\& on a
3789 machine acting as a router, if we want to limit traffic from
3790 local clients on 192.168.2.0/24 we do:
3792 .Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out"
3793 .Dl "ipfw pipe 1 config bw 300Kbit/s queue 50KBytes"
3795 note that we use the
3797 modifier so that the rule is not used twice.
3798 Remember in fact that
3800 rules are checked both on incoming and outgoing packets.
3802 Should we want to simulate a bidirectional link with bandwidth
3803 limitations, the correct way is the following:
3805 .Dl "ipfw add pipe 1 ip from any to any out"
3806 .Dl "ipfw add pipe 2 ip from any to any in"
3807 .Dl "ipfw pipe 1 config bw 64Kbit/s queue 10Kbytes"
3808 .Dl "ipfw pipe 2 config bw 64Kbit/s queue 10Kbytes"
3810 The above can be very useful, e.g.\& if you want to see how
3811 your fancy Web page will look for a residential user who
3812 is connected only through a slow link.
3813 You should not use only one pipe for both directions, unless
3814 you want to simulate a half-duplex medium (e.g.\& AppleTalk,
3816 It is not necessary that both pipes have the same configuration,
3817 so we can also simulate asymmetric links.
3819 Should we want to verify network performance with the RED queue
3820 management algorithm:
3822 .Dl "ipfw add pipe 1 ip from any to any"
3823 .Dl "ipfw pipe 1 config bw 500Kbit/s queue 100 red 0.002/30/80/0.1"
3825 Another typical application of the traffic shaper is to
3826 introduce some delay in the communication.
3827 This can significantly affect applications which do a lot of Remote
3828 Procedure Calls, and where the round-trip-time of the
3829 connection often becomes a limiting factor much more than
3832 .Dl "ipfw add pipe 1 ip from any to any out"
3833 .Dl "ipfw add pipe 2 ip from any to any in"
3834 .Dl "ipfw pipe 1 config delay 250ms bw 1Mbit/s"
3835 .Dl "ipfw pipe 2 config delay 250ms bw 1Mbit/s"
3837 Per-flow queueing can be useful for a variety of purposes.
3838 A very simple one is counting traffic:
3840 .Dl "ipfw add pipe 1 tcp from any to any"
3841 .Dl "ipfw add pipe 1 udp from any to any"
3842 .Dl "ipfw add pipe 1 ip from any to any"
3843 .Dl "ipfw pipe 1 config mask all"
3845 The above set of rules will create queues (and collect
3846 statistics) for all traffic.
3847 Because the pipes have no limitations, the only effect is
3848 collecting statistics.
3849 Note that we need 3 rules, not just the last one, because
3852 tries to match IP packets it will not consider ports, so we
3853 would not see connections on separate ports as different
3856 A more sophisticated example is limiting the outbound traffic
3857 on a net with per-host limits, rather than per-network limits:
3859 .Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out"
3860 .Dl "ipfw add pipe 2 ip from any to 192.168.2.0/24 in"
3861 .Dl "ipfw pipe 1 config mask src-ip 0x000000ff bw 200Kbit/s queue 20Kbytes"
3862 .Dl "ipfw pipe 2 config mask dst-ip 0x000000ff bw 200Kbit/s queue 20Kbytes"
3864 In the following example, we need to create several traffic bandwidth
3865 classes and we need different hosts/networks to fall into different classes.
3866 We create one pipe for each class and configure them accordingly.
3867 Then we create a single table and fill it with IP subnets and addresses.
3868 For each subnet/host we set the argument equal to the number of the pipe
3870 Then we classify traffic using a single rule:
3872 .Dl "ipfw pipe 1 config bw 1000Kbyte/s"
3873 .Dl "ipfw pipe 4 config bw 4000Kbyte/s"
3875 .Dl "ipfw table T1 create type addr"
3876 .Dl "ipfw table T1 add 192.168.2.0/24 1"
3877 .Dl "ipfw table T1 add 192.168.0.0/27 4"
3878 .Dl "ipfw table T1 add 192.168.0.2 1"
3880 .Dl "ipfw add pipe tablearg ip from 'table(T1)' to any"
3884 action, the table entries may include hostnames and IP addresses.
3886 .Dl "ipfw table T2 create type addr ftype ip"
3887 .Dl "ipfw table T2 add 192.168.2.0/24 10.23.2.1"
3888 .Dl "ipfw table T21 add 192.168.0.0/27 router1.dmz"
3890 .Dl "ipfw add 100 fwd tablearg ip from any to table(1)"
3892 In the following example per-interface firewall is created:
3894 .Dl "ipfw table IN create type iface valtype skipto,fib"
3895 .Dl "ipfw table IN add vlan20 12000,12"
3896 .Dl "ipfw table IN add vlan30 13000,13"
3897 .Dl "ipfw table OUT create type iface valtype skipto"
3898 .Dl "ipfw table OUT add vlan20 22000"
3899 .Dl "ipfw table OUT add vlan30 23000"
3901 .Dl "ipfw add 100 setfib tablearg ip from any to any recv 'table(IN)' in"
3902 .Dl "ipfw add 200 skipto tablearg ip from any to any recv 'table(IN)' in"
3903 .Dl "ipfw add 300 skipto tablearg ip from any to any xmit 'table(OUT)' out"
3905 The following example illustrate usage of flow tables:
3907 .Dl "ipfw table fl create type flow:flow:src-ip,proto,dst-ip,dst-port"
3908 .Dl "ipfw table fl add 2a02:6b8:77::88,tcp,2a02:6b8:77::99,80 11"
3909 .Dl "ipfw table fl add 10.0.0.1,udp,10.0.0.2,53 12"
3911 .Dl "ipfw add 100 allow ip from any to any flow 'table(fl,11)' recv ix0"
3913 To add a set of rules atomically, e.g.\& set 18:
3915 .Dl "ipfw set disable 18"
3916 .Dl "ipfw add NN set 18 ... # repeat as needed"
3917 .Dl "ipfw set enable 18"
3919 To delete a set of rules atomically the command is simply:
3921 .Dl "ipfw delete set 18"
3923 To test a ruleset and disable it and regain control if something goes wrong:
3925 .Dl "ipfw set disable 18"
3926 .Dl "ipfw add NN set 18 ... # repeat as needed"
3927 .Dl "ipfw set enable 18; echo done; sleep 30 && ipfw set disable 18"
3929 Here if everything goes well, you press control-C before the "sleep"
3930 terminates, and your ruleset will be left active.
3931 Otherwise, e.g.\& if
3932 you cannot access your box, the ruleset will be disabled after
3933 the sleep terminates thus restoring the previous situation.
3935 To show rules of the specific set:
3937 .Dl "ipfw set 18 show"
3939 To show rules of the disabled set:
3941 .Dl "ipfw -S set 18 show"
3943 To clear a specific rule counters of the specific set:
3945 .Dl "ipfw set 18 zero NN"
3947 To delete a specific rule of the specific set:
3949 .Dl "ipfw set 18 delete NN"
3950 .Ss NAT, REDIRECT AND LSNAT
3951 First redirect all the traffic to nat instance 123:
3953 .Dl "ipfw add nat 123 all from any to any"
3955 Then to configure nat instance 123 to alias all the outgoing traffic with ip
3956 192.168.0.123, blocking all incoming connections, trying to keep
3957 same ports on both sides, clearing aliasing table on address change
3958 and keeping a log of traffic/link statistics:
3960 .Dl "ipfw nat 123 config ip 192.168.0.123 log deny_in reset same_ports"
3962 Or to change address of instance 123, aliasing table will be cleared (see
3965 .Dl "ipfw nat 123 config ip 10.0.0.1"
3967 To see configuration of nat instance 123:
3969 .Dl "ipfw nat 123 show config"
3971 To show logs of all the instances in range 111-999:
3973 .Dl "ipfw nat 111-999 show"
3975 To see configurations of all instances:
3977 .Dl "ipfw nat show config"
3979 Or a redirect rule with mixed modes could looks like:
3981 .Dl "ipfw nat 123 config redirect_addr 10.0.0.1 10.0.0.66"
3982 .Dl " redirect_port tcp 192.168.0.1:80 500"
3983 .Dl " redirect_proto udp 192.168.1.43 192.168.1.1"
3984 .Dl " redirect_addr 192.168.0.10,192.168.0.11"
3985 .Dl " 10.0.0.100 # LSNAT"
3986 .Dl " redirect_port tcp 192.168.0.1:80,192.168.0.10:22"
3989 or it could be split in:
3991 .Dl "ipfw nat 1 config redirect_addr 10.0.0.1 10.0.0.66"
3992 .Dl "ipfw nat 2 config redirect_port tcp 192.168.0.1:80 500"
3993 .Dl "ipfw nat 3 config redirect_proto udp 192.168.1.43 192.168.1.1"
3994 .Dl "ipfw nat 4 config redirect_addr 192.168.0.10,192.168.0.11,192.168.0.12"
3996 .Dl "ipfw nat 5 config redirect_port tcp"
3997 .Dl " 192.168.0.1:80,192.168.0.10:22,192.168.0.20:25 500"
3999 Sometimes you may want to mix NAT and dynamic rules. It could be achived with
4003 options. Problem is, you need to create dynamic rule before NAT and check it
4004 after NAT actions (or vice versa) to have consistent addresses and ports.
4007 option will trigger activation of existing dynamic state, and action of such
4008 rule will be performed as soon as rule is matched. In case of NAT and
4010 rule packet need to be passed to NAT, not allowed as soon is possible.
4012 There is example of set of rules to achive this. Bear in mind that this
4013 is exmaple only and it is not very usefult by itself.
4015 On way out, after all checks place this rules:
4017 .Dl "ipfw add allow record-state skip-action"
4018 .Dl "ipfw add nat 1"
4020 And on way in there should be something like this:
4022 .Dl "ipfw add nat 1"
4023 .Dl "ipfw add check-state"
4025 Please note, that first rule on way out doesn't allow packet and doesn't
4026 execute existing dynamic rules. All it does, create new dynamic rule with
4028 action, if it is not created yet. Later, this dynamic rule is used on way
4052 utility first appeared in
4057 Stateful extensions were introduced in
4060 was introduced in Summer 2002.
4062 .An Ugen J. S. Antsilevich ,
4063 .An Poul-Henning Kamp ,
4069 API based upon code written by
4073 Dummynet has been introduced by Luigi Rizzo in 1997-1998.
4075 Some early work (1999-2000) on the
4077 traffic shaper supported by Akamba Corp.
4079 The ipfw core (ipfw2) has been completely redesigned and
4080 reimplemented by Luigi Rizzo in summer 2002.
4083 options have been added by various developer over the years.
4086 In-kernel NAT support written by
4087 .An Paolo Pisati Aq Mt piso@FreeBSD.org
4088 as part of a Summer of Code 2005 project.
4092 support has been developed by
4093 .An The Centre for Advanced Internet Architectures (CAIA) Aq http://www.caia.swin.edu.au .
4094 The primary developers and maintainers are David Hayes and Jason But.
4095 For further information visit:
4096 .Aq http://www.caia.swin.edu.au/urp/SONATA
4098 Delay profiles have been developed by Alessandro Cerri and
4099 Luigi Rizzo, supported by the
4100 European Commission within Projects Onelab and Onelab2.
4102 The syntax has grown over the years and sometimes it might be confusing.
4103 Unfortunately, backward compatibility prevents cleaning up mistakes
4104 made in the definition of the syntax.
4108 Misconfiguring the firewall can put your computer in an unusable state,
4109 possibly shutting down network services and requiring console access to
4110 regain control of it.
4112 Incoming packet fragments diverted by
4114 are reassembled before delivery to the socket.
4115 The action used on those packet is the one from the
4116 rule which matches the first fragment of the packet.
4118 Packets diverted to userland, and then reinserted by a userland process
4119 may lose various packet attributes.
4120 The packet source interface name
4121 will be preserved if it is shorter than 8 bytes and the userland process
4122 saves and reuses the sockaddr_in
4125 otherwise, it may be lost.
4126 If a packet is reinserted in this manner, later rules may be incorrectly
4127 applied, making the order of
4129 rules in the rule sequence very important.
4131 Dummynet drops all packets with IPv6 link-local addresses.
4137 may not behave as expected.
4138 In particular, incoming SYN packets may
4139 have no uid or gid associated with them since they do not yet belong
4140 to a TCP connection, and the uid/gid associated with a packet may not
4141 be as expected if the associated process calls
4143 or similar system calls.
4145 Rule syntax is subject to the command line environment and some patterns
4146 may need to be escaped with the backslash character
4147 or quoted appropriately.
4149 Due to the architecture of
4151 ipfw nat is not compatible with the TCP segmentation offloading (TSO).
4152 Thus, to reliably nat your network traffic, please disable TSO
4156 ICMP error messages are not implicitly matched by dynamic rules
4157 for the respective conversations.
4158 To avoid failures of network error detection and path MTU discovery,
4159 ICMP error messages may need to be allowed explicitly through static
4166 actions may lead to confusing behaviour if ruleset has mistakes,
4167 and/or interaction with other subsystems (netgraph, dummynet, etc.) is used.
4168 One possible case for this is packet leaving
4170 in subroutine on the input pass, while later on output encountering unpaired
4173 As the call stack is kept intact after input pass, packet will suddenly
4174 return to the rule number used on input pass, not on output one.
4175 Order of processing should be checked carefully to avoid such mistakes.