1 .\" $OpenBSD: pfctl.8,v 1.138 2008/06/10 20:55:02 mcbride Exp $
3 .\" Copyright (c) 2001 Kjell Wooding. All rights reserved.
5 .\" Redistribution and use in source and binary forms, with or without
6 .\" modification, are permitted provided that the following conditions
8 .\" 1. Redistributions of source code must retain the above copyright
9 .\" notice, this list of conditions and the following disclaimer.
10 .\" 2. Redistributions in binary form must reproduce the above copyright
11 .\" notice, this list of conditions and the following disclaimer in the
12 .\" documentation and/or other materials provided with the distribution.
13 .\" 3. The name of the author may not be used to endorse or promote products
14 .\" derived from this software without specific prior written permission.
16 .\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
17 .\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
18 .\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
19 .\" IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
20 .\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
21 .\" NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
22 .\" DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
23 .\" THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
24 .\" (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
25 .\" THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
34 .Nd control the packet filter (PF) device
38 .Op Fl AdeghmNnOPqRrvz
40 .Oo Fl D Ar macro Ns =
45 .Op Fl K Ar host | network
48 .Ar host | network | label | id
63 utility communicates with the packet filter device using the
64 ioctl interface described in
66 It allows ruleset and parameter configuration and retrieval of status
67 information from the packet filter.
69 Packet filtering restricts the types of packets that pass through
70 network interfaces entering or leaving the host based on filter
73 The packet filter can also replace addresses and ports of packets.
74 Replacing source addresses and ports of outgoing packets is called
75 NAT (Network Address Translation) and is used to connect an internal
76 network (usually reserved address space) to an external one (the
77 Internet) by making all connections to external hosts appear to
78 come from the gateway.
79 Replacing destination addresses and ports of incoming packets
80 is used to redirect connections to different hosts and/or ports.
81 A combination of both translations, bidirectional NAT, is also
83 Translation rules are described in
92 the rule file specified with the variable
94 is loaded automatically by the
96 scripts and the packet filter is enabled.
98 The packet filter does not itself forward packets between interfaces.
99 Forwarding can be enabled by setting the
102 .Em net.inet.ip.forwarding
104 .Em net.inet6.ip6.forwarding
106 Set them permanently in
111 utility provides several commands.
112 The options are as follows:
115 Load only the queue rules present in the rule file.
116 Other rules and options are ignored.
123 only to the rules in the specified
125 In addition to the main ruleset,
127 can load and manipulate additional rulesets by name,
129 The main ruleset is the default anchor.
131 Anchors are referenced by name and may be nested,
132 with the various components of the anchor path separated by
134 characters, similar to how file system hierarchies are laid out.
135 The last component of the anchor path is where ruleset operations are
140 rules from the main ruleset is described in
143 For example, the following will show all filter rules (see the
145 flag below) inside the anchor
146 .Dq authpf/smith(1234) ,
147 which would have been created for user
152 .Bd -literal -offset indent
153 # pfctl -a "authpf/smith(1234)" -s rules
156 Private tables can also be put inside anchors, either by having table
159 file that is loaded in the anchor, or by using regular table commands, as in:
160 .Bd -literal -offset indent
161 # pfctl -a foo/bar -t mytable -T add 1.2.3.4 5.6.7.8
164 When a rule referring to a table is loaded in an anchor, the rule will use the
165 private table if one is defined, and then fall back to the table defined in the
166 main ruleset, if there is one.
167 This is similar to C rules for variable scope.
168 It is possible to create distinct tables with the same name in the global
169 ruleset and in an anchor, but this is often bad design and a warning will be
172 By default, recursive inline printing of anchors applies only to unnamed
173 anchors specified inline in the ruleset.
174 If the anchor name is terminated with a
178 flag will recursively print all anchors in a brace delimited block.
179 For example the following will print the
182 .Bd -literal -offset indent
183 # pfctl -a 'authpf/*' -sr
186 To print the main ruleset recursively, specify only
189 .Bd -literal -offset indent
192 .It Fl D Ar macro Ns = Ns Ar value
198 Overrides the definition of
202 Disable the packet filter.
204 Enable the packet filter.
206 Flush the filter parameters specified by
208 (may be abbreviated):
210 .Bl -tag -width xxxxxxxxxxxx -compact
214 Flush the queue rules.
216 Flush the filter rules.
218 Flush the state table (NAT and filter).
220 Flush the source tracking table.
222 Flush the filter information (statistics that are not bound to rules).
226 Flush the passive operating system fingerprints.
228 Flush all of the above.
231 Load the rules contained in
235 may contain macros, tables, options, and normalization, queueing,
236 translation, and filtering rules.
237 With the exception of macros and tables, the statements must appear in that
240 Include output helpful for debugging.
243 .It Fl i Ar interface
244 Restrict the operation to the given
246 .It Fl K Ar host | network
247 Kill all of the source tracking entries originating from the specified
255 option may be specified, which will kill all the source tracking
256 entries from the first host/network to the second.
259 .Ar host | network | label | id
261 Kill all of the state entries matching the specified
268 For example, to kill all of the state entries originating from
277 option may be specified, which will kill all the state entries
278 from the first host/network to the second.
279 To kill all of the state entries from
284 .Dl # pfctl -k host1 -k host2
286 To kill all states originating from 192.168.1.0/24 to 172.16.0.0/16:
288 .Dl # pfctl -k 192.168.1.0/24 -k 172.16.0.0/16
290 A network prefix length of 0 can be used as a wildcard.
291 To kill all states with the target
294 .Dl # pfctl -k 0.0.0.0/0 -k host2
296 It is also possible to kill states by rule label or state ID.
297 In this mode the first
299 argument is used to specify the type
300 of the second argument.
301 The following command would kill all states that have been created
302 from rules carrying the label
305 .Dl # pfctl -k label -k foobar
307 To kill one specific state by its unique state ID
308 (as shown by pfctl -s state -vv),
311 modifier and as a second argument the state ID and optional creator ID.
312 To kill a state with ID 4823e84500000003 use:
314 .Dl # pfctl -k id -k 4823e84500000003
316 To kill a state with ID 4823e84500000018 created from a backup
317 firewall with hostid 00000002 use:
319 .Dl # pfctl -k id -k 4823e84500000018/2
321 Merge in explicitly given options without resetting those
323 Allows single options to be modified without disturbing the others:
324 .Bd -literal -offset indent
325 # echo "set loginterface fxp0" | pfctl -mf -
328 Load only the NAT rules present in the rule file.
329 Other rules and options are ignored.
331 Do not actually load rules, just parse them.
333 Load only the options present in the rule file.
334 Other rules and options are ignored.
336 Control the ruleset optimizer, overriding any rule file settings.
338 .Bl -tag -width xxxxxxxxxxxx -compact
340 Disable the ruleset optimizer.
342 Enable basic ruleset optimizations.
343 This is the default behaviour.
345 Enable basic ruleset optimizations with profiling.
347 For further information on the ruleset optimizer, see
350 Do not perform service name lookup for port specific rules,
351 instead display the ports numerically.
355 instead of the default
358 Only print errors and warnings.
360 Load only the filter rules present in the rule file.
361 Other rules and options are ignored.
363 Perform reverse DNS lookups on states when displaying them.
365 Show the filter parameters specified by
367 (may be abbreviated):
369 .Bl -tag -width xxxxxxxxxxxxx -compact
371 Show the currently loaded NAT rules.
373 Show the currently loaded queue rules.
374 When used together with
376 per-queue statistics are also shown.
377 When used together with
380 will loop and show updated queue statistics every five seconds, including
381 measured bandwidth and packets per second.
383 Show the currently loaded filter rules.
384 When used together with
386 the per-rule statistics (number of evaluations,
387 packets and bytes) are also shown.
390 optimization done automatically by the kernel
391 will skip evaluation of rules where possible.
392 Packets passed statefully are counted in the rule that created the state
393 (even though the rule is not evaluated more than once for the entire
396 Show the currently loaded anchors directly attached to the main ruleset.
399 is specified as well, the anchors loaded directly below the given
404 is specified, all anchors attached under the target anchor will be
405 displayed recursively.
407 Show the contents of the state table.
409 Show the contents of the source tracking table.
411 Show filter information (statistics and counters).
412 When used together with
414 source tracking statistics are also shown.
416 Show the running status and provide a non-zero exit status when disabled.
418 Show per-rule statistics (label, evaluations, packets total, bytes total,
419 packets in, bytes in, packets out, bytes out, state creations) of
420 filter rules with labels, useful for accounting.
422 Show the current global timeouts.
424 Show the current pool memory hard limits.
426 Show the list of tables.
428 Show the list of operating system fingerprints.
429 .It Fl s Cm Interfaces
430 Show the list of interfaces and interface drivers available to PF.
431 When used together with
433 it additionally lists which interfaces have skip rules activated.
434 When used together with
436 interface statistics are also shown.
438 can be used to select an interface or a group of interfaces.
440 Show all of the above, except for the lists of interfaces and operating
443 .It Fl T Ar command Op Ar address ...
446 (may be abbreviated) to apply to the table.
449 .Bl -tag -width xxxxxxxxxxxx -compact
453 Flush all addresses of a table.
455 Add one or more addresses in a table.
456 Automatically create a nonexisting table.
458 Delete one or more addresses from a table.
459 .It Fl T Cm expire Ar number
460 Delete addresses which had their statistics cleared more than
463 For entries which have never had their statistics cleared,
465 refers to the time they were added to the table.
467 Replace the addresses of the table.
468 Automatically create a nonexisting table.
470 Show the content (addresses) of a table.
472 Test if the given addresses match a table.
474 Clear all the statistics of a table.
476 Load only the table definitions from
478 This is used in conjunction with the
481 .Bd -literal -offset indent
482 # pfctl -Tl -f pf.conf
492 commands, the list of addresses can be specified either directly on the command
493 line and/or in an unformatted text file, using the
496 Comments starting with a
498 are allowed in the text file.
499 With these commands, the
501 flag can also be used once or twice, in which case
504 detailed result of the operation for each individual address, prefixed by
505 one of the following letters:
507 .Bl -tag -width XXX -compact
509 The address/network has been added.
511 The address/network has been changed (negated).
513 The address/network has been deleted.
521 The address/network is duplicated and therefore ignored.
523 The address/network cannot be added/deleted due to conflicting
527 The address/network has been cleared (statistics).
530 Each table can maintain a set of counters that can be retrieved using the
534 For example, the following commands define a wide open firewall which will keep
535 track of packets going to or coming from the
538 The following commands configure the firewall and send 10 pings to the FTP
540 .Bd -literal -offset indent
541 # printf "table <test> counters { ftp.openbsd.org }\en \e
542 pass out to <test>\en" | pfctl -f-
543 # ping -qc10 ftp.openbsd.org
546 We can now use the table
548 command to output, for each address and packet direction, the number of packets
549 and bytes that are being passed or blocked by rules referencing the table.
550 The time at which the current accounting started is also shown with the
553 .Bd -literal -offset indent
554 # pfctl -t test -vTshow
556 Cleared: Thu Feb 13 18:55:18 2003
557 In/Block: [ Packets: 0 Bytes: 0 ]
558 In/Pass: [ Packets: 10 Bytes: 840 ]
559 Out/Block: [ Packets: 0 Bytes: 0 ]
560 Out/Pass: [ Packets: 10 Bytes: 840 ]
563 Similarly, it is possible to view global information about the tables
566 modifier twice and the
570 This will display the number of addresses on each table,
571 the number of rules which reference the table, and the global
572 packet statistics for the whole table:
573 .Bd -literal -offset indent
577 Cleared: Thu Feb 13 18:55:18 2003
578 References: [ Anchors: 0 Rules: 1 ]
579 Evaluations: [ NoMatch: 3496 Match: 1 ]
580 In/Block: [ Packets: 0 Bytes: 0 ]
581 In/Pass: [ Packets: 10 Bytes: 840 ]
582 In/XPass: [ Packets: 0 Bytes: 0 ]
583 Out/Block: [ Packets: 0 Bytes: 0 ]
584 Out/Pass: [ Packets: 10 Bytes: 840 ]
585 Out/XPass: [ Packets: 0 Bytes: 0 ]
588 As we can see here, only one packet \- the initial ping request \- matched the
589 table, but all packets passing as the result of the state are correctly
591 Reloading the table(s) or ruleset will not affect packet accounting in any way.
594 counters are incremented instead of the
598 packet is passed but does not match the table anymore.
599 This will happen in our example if someone flushes the table while the
603 When used with a single
606 will only display the first line containing the table flags and name.
607 The flags are defined as follows:
609 .Bl -tag -width XXX -compact
611 For constant tables, which cannot be altered outside
614 For persistent tables, which do not get automatically killed when no rules
617 For tables which are part of the
620 Tables without this flag do not really exist, cannot contain addresses, and are
625 For tables which are part of the
628 This flag can only be witnessed briefly during the loading of
631 For tables which are referenced (used) by rules.
633 This flag is set when a table in the main ruleset is hidden by one or more
634 tables of the same name from anchors attached below it.
636 This flag is set when per-address counters are enabled on the table.
639 Specify the name of the table.
641 Produce more verbose output.
644 will produce even more verbose output including ruleset warnings.
645 See the previous section for its effect on table commands.
649 (may be abbreviated) to one of the following:
651 .Bl -tag -width xxxxxxxxxxxx -compact
653 Do not generate debug messages.
655 Generate debug messages only for serious errors.
657 Generate debug messages for various errors.
659 Generate debug messages for common conditions.
662 Clear per-rule statistics.
665 .Bl -tag -width "/etc/pf.conf" -compact
667 Packet filter rules file.
669 Passive operating system fingerprint database.
687 filter mechanism appeared in
689 They first appeared in
691 ported from the version in