2 .\" @(#) $Header: /tcpdump/master/tcpdump/tcpdump.1.in,v 1.2 2008-11-09 23:35:03 mcr Exp $ (LBL)
4 .\" $NetBSD: tcpdump.8,v 1.9 2003/03/31 00:18:17 perry Exp $
6 .\" Copyright (c) 1987, 1988, 1989, 1990, 1991, 1992, 1994, 1995, 1996, 1997
7 .\" The Regents of the University of California. All rights reserved.
8 .\" All rights reserved.
10 .\" Redistribution and use in source and binary forms, with or without
11 .\" modification, are permitted provided that: (1) source code distributions
12 .\" retain the above copyright notice and this paragraph in its entirety, (2)
13 .\" distributions including binary code include the above copyright notice and
14 .\" this paragraph in its entirety in the documentation or other materials
15 .\" provided with the distribution, and (3) all advertising materials mentioning
16 .\" features or use of this software display the following acknowledgement:
17 .\" ``This product includes software developed by the University of California,
18 .\" Lawrence Berkeley Laboratory and its contributors.'' Neither the name of
19 .\" the University nor the names of its contributors may be used to endorse
20 .\" or promote products derived from this software without specific prior
21 .\" written permission.
22 .\" THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED
23 .\" WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
24 .\" MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
26 .TH TCPDUMP 1 "05 March 2009"
28 tcpdump \- dump traffic on a network
33 .B \-AbdDefIKlLnNOpqRStuUvxX
95 .I spi@ipaddr algo:secret,...
105 .I postrotate-command
119 \fITcpdump\fP prints out a description of the contents of packets on a
120 network interface that match the boolean \fIexpression\fP. It can also
123 flag, which causes it to save the packet data to a file for later
124 analysis, and/or with the
126 flag, which causes it to read from a saved packet file rather than to
127 read packets from a network interface. In all cases, only packets that
134 will, if not run with the
136 flag, continue capturing packets until it is interrupted by a SIGINT
137 signal (generated, for example, by typing your interrupt character,
138 typically control-C) or a SIGTERM signal (typically generated with the
140 command); if run with the
142 flag, it will capture packets until it is interrupted by a SIGINT or
143 SIGTERM signal or the specified number of packets have been processed.
147 finishes capturing packets, it will report counts of:
149 packets ``captured'' (this is the number of packets that
151 has received and processed);
153 packets ``received by filter'' (the meaning of this depends on the OS on
156 and possibly on the way the OS was configured - if a filter was
157 specified on the command line, on some OSes it counts packets regardless
158 of whether they were matched by the filter expression and, even if they
159 were matched by the filter expression, regardless of whether
161 has read and processed them yet, on other OSes it counts only packets that were
162 matched by the filter expression regardless of whether
164 has read and processed them yet, and on other OSes it counts only
165 packets that were matched by the filter expression and were processed by
168 packets ``dropped by kernel'' (this is the number of packets that were
169 dropped, due to a lack of buffer space, by the packet capture mechanism
172 is running, if the OS reports that information to applications; if not,
173 it will be reported as 0).
175 On platforms that support the SIGINFO signal, such as most BSDs
176 (including Mac OS X) and Digital/Tru64 UNIX, it will report those counts
177 when it receives a SIGINFO signal (generated, for example, by typing
178 your ``status'' character, typically control-T, although on some
179 platforms, such as Mac OS X, the ``status'' character is not set by
180 default, so you must set it with
182 in order to use it) and will continue capturing packets.
184 Reading packets from a network interface may require that you have
185 special privileges; see the
187 man page for details. Reading a saved packet file doesn't require
192 Print each packet (minus its link level header) in ASCII. Handy for
196 Print the AS number in BGP packets in ASDOT notation rather than ASPLAIN
200 Set the operating system capture buffer size to \fIbuffer_size\fP.
203 Exit after receiving \fIcount\fP packets.
206 Before writing a raw packet to a savefile, check whether the file is
207 currently larger than \fIfile_size\fP and, if so, close the current
208 savefile and open a new one. Savefiles after the first savefile will
209 have the name specified with the
211 flag, with a number after it, starting at 1 and continuing upward.
212 The units of \fIfile_size\fP are millions of bytes (1,000,000 bytes,
213 not 1,048,576 bytes).
216 Dump the compiled packet-matching code in a human readable form to
217 standard output and stop.
220 Dump packet-matching code as a
225 Dump packet-matching code as decimal numbers (preceded with a count).
228 Print the list of the network interfaces available on the system and on
231 can capture packets. For each network interface, a number and an
232 interface name, possibly followed by a text description of the
233 interface, is printed. The interface name or the number can be supplied
236 flag to specify an interface on which to capture.
238 This can be useful on systems that don't have a command to list them
239 (e.g., Windows systems, or UNIX systems lacking
240 .BR "ifconfig \-a" );
241 the number can be useful on Windows 2000 and later systems, where the
242 interface name is a somewhat complex string.
246 flag will not be supported if
248 was built with an older version of
251 .B pcap_findalldevs()
255 Print the link-level header on each dump line.
258 Use \fIspi@ipaddr algo:secret\fP for decrypting IPsec ESP packets that
259 are addressed to \fIaddr\fP and contain Security Parameter Index value
260 \fIspi\fP. This combination may be repeated with comma or newline seperation.
262 Note that setting the secret for IPv4 ESP packets is supported at this time.
269 \fBcast128-cbc\fP, or
271 The default is \fBdes-cbc\fP.
272 The ability to decrypt packets is only present if \fItcpdump\fP was compiled
273 with cryptography enabled.
275 \fIsecret\fP is the ASCII text for ESP secret key.
276 If preceeded by 0x, then a hex value will be read.
278 The option assumes RFC2406 ESP, not RFC1827 ESP.
279 The option is only for debugging purposes, and
280 the use of this option with a true `secret' key is discouraged.
281 By presenting IPsec secret key onto command line
282 you make it visible to others, via
286 In addition to the above syntax, the syntax \fIfile name\fP may be used
287 to have tcpdump read the provided file in. The file is opened upon
288 receiving the first ESP packet, so any special permissions that tcpdump
289 may have been given should already have been given up.
292 Print `foreign' IPv4 addresses numerically rather than symbolically
293 (this option is intended to get around serious brain damage in
294 Sun's NIS server \(em usually it hangs forever translating non-local
297 The test for `foreign' IPv4 addresses is done using the IPv4 address and
298 netmask of the interface on which capture is being done. If that
299 address or netmask are not available, available, either because the
300 interface on which capture is being done has no address or netmask or
301 because the capture is being done on the Linux "any" interface, which
302 can capture on more than one interface, this option will not work
306 Use \fIfile\fP as input for the filter expression.
307 An additional expression given on the command line is ignored.
310 If specified, rotates the dump file specified with the
312 option every \fIrotate_seconds\fP seconds.
313 Savefiles will have the name specified by
315 which should include a time format as defined by
317 If no time format is specified, each new file will overwrite the previous.
319 If used in conjunction with the
321 option, filenames will take the form of `\fIfile\fP<count>'.
324 Listen on \fIinterface\fP.
325 If unspecified, \fItcpdump\fP searches the system interface list for the
326 lowest numbered, configured up interface (excluding loopback).
327 Ties are broken by choosing the earliest match.
329 On Linux systems with 2.2 or later kernels, an
331 argument of ``any'' can be used to capture packets from all interfaces.
332 Note that captures on the ``any'' device will not be done in promiscuous
337 flag is supported, an interface number as printed by that flag can be
343 Put the interface in "monitor mode"; this is supported only on IEEE
344 802.11 Wi-Fi interfaces, and supported only on some operating systems.
346 Note that in monitor mode the adapter might disassociate from the
347 network with which it's associated, so that you will not be able to use
348 any wireless networks with that adapter. This could prevent accessing
349 files on a network server, or resolving host names or network addresses,
350 if you are capturing in monitor mode and are not connected to another
351 network with another adapter.
353 This flag will affect the output of the
357 isn't specified, only those link-layer types available when not in
358 monitor mode will be shown; if
360 is specified, only those link-layer types available when in monitor mode
364 Don't attempt to verify IP, TCP, or UDP checksums. This is useful for
365 interfaces that perform some or all of those checksum calculation in
366 hardware; otherwise, all outgoing TCP checksums will be flagged as bad.
369 Make stdout line buffered.
370 Useful if you want to see the data
374 ``tcpdump\ \ \-l\ \ |\ \ tee dat'' or
375 ``tcpdump\ \ \-l \ \ > dat\ \ &\ \ tail\ \ \-f\ \ dat''.
378 List the known data link types for the interface, in the specified mode,
379 and exit. The list of known data link types may be dependent on the
380 specified mode; for example, on some platforms, a Wi-Fi interface might
381 support one set of data link types when not in monitor mode (for
382 example, it might support only fake Ethernet headers, or might support
383 802.11 headers but not support 802.11 headers with radio information)
384 and another set of data link types when in monitor mode (for example, it
385 might support 802.11 headers, or 802.11 headers with radio information,
386 only in monitor mode).
389 Load SMI MIB module definitions from file \fImodule\fR.
391 can be used several times to load several MIB modules into \fItcpdump\fP.
394 Use \fIsecret\fP as a shared secret for validating the digests found in
395 TCP segments with the TCP-MD5 option (RFC 2385), if present.
398 Don't convert addresses (i.e., host addresses, port numbers, etc.) to names.
401 Don't print domain name qualification of host names.
403 if you give this flag then \fItcpdump\fP will print ``nic''
404 instead of ``nic.ddn.mil''.
407 Do not run the packet-matching code optimizer.
409 if you suspect a bug in the optimizer.
412 \fIDon't\fP put the interface
413 into promiscuous mode.
414 Note that the interface might be in promiscuous
415 mode for some other reason; hence, `-p' cannot be used as an abbreviation for
416 `ether host {local-hw-addr} or ether broadcast'.
419 Quick (quiet?) output.
420 Print less protocol information so output
424 Assume ESP/AH packets to be based on old specification (RFC1825 to RFC1829).
425 If specified, \fItcpdump\fP will not print replay prevention field.
426 Since there is no protocol version field in ESP/AH specification,
427 \fItcpdump\fP cannot deduce the version of ESP/AH protocol.
430 Read packets from \fIfile\fR (which was created with the
433 Standard input is used if \fIfile\fR is ``-''.
436 Print absolute, rather than relative, TCP sequence numbers.
439 Snarf \fIsnaplen\fP bytes of data from each packet rather than the
440 default of 65535 bytes.
441 Packets truncated because of a limited snapshot
442 are indicated in the output with ``[|\fIproto\fP]'', where \fIproto\fP
443 is the name of the protocol level at which the truncation has occurred.
444 Note that taking larger snapshots both increases
445 the amount of time it takes to process packets and, effectively,
446 decreases the amount of packet buffering.
447 This may cause packets to be
449 You should limit \fIsnaplen\fP to the smallest number that will
450 capture the protocol information you're interested in.
452 \fIsnaplen\fP to 0 sets it to the default of 65535,
453 for backwards compatibility with recent older versions of
457 Force packets selected by "\fIexpression\fP" to be interpreted the
458 specified \fItype\fR.
459 Currently known types are
460 \fBaodv\fR (Ad-hoc On-demand Distance Vector protocol),
461 \fBcnfp\fR (Cisco NetFlow protocol),
462 \fBrpc\fR (Remote Procedure Call),
463 \fBrtp\fR (Real-Time Applications protocol),
464 \fBrtcp\fR (Real-Time Applications control protocol),
465 \fBsnmp\fR (Simple Network Management Protocol),
466 \fBtftp\fR (Trivial File Transfer Protocol),
467 \fBvat\fR (Visual Audio Tool),
469 \fBwb\fR (distributed White Board).
472 \fIDon't\fP print a timestamp on each dump line.
475 Print an unformatted timestamp on each dump line.
478 Print a delta (micro-second resolution) between current and previous line
482 Print a timestamp in default format proceeded by date on each dump line.
485 Print a delta (micro-second resolution) between current and first line
489 Print undecoded NFS handles.
492 Make output saved via the
494 option ``packet-buffered''; i.e., as each packet is saved, it will be
495 written to the output file, rather than being written only when the
500 flag will not be supported if
502 was built with an older version of
509 When parsing and printing, produce (slightly more) verbose output.
510 For example, the time to live,
511 identification, total length and options in an IP packet are printed.
512 Also enables additional packet integrity checks such as verifying the
513 IP and ICMP header checksum.
515 When writing to a file with the
517 option, report, every 10 seconds, the number of packets captured.
520 Even more verbose output.
521 For example, additional fields are
522 printed from NFS reply packets, and SMB packets are fully decoded.
525 Even more verbose output.
527 telnet \fBSB\fP ... \fBSE\fP options
531 Telnet options are printed in hex as well.
534 Write the raw packets to \fIfile\fR rather than parsing and printing
536 They can later be printed with the \-r option.
537 Standard output is used if \fIfile\fR is ``-''.
539 .BR pcap-savefile (5)
540 for a description of the file format.
543 Used in conjunction with the
545 option, this will limit the number
546 of files created to the specified number, and begin overwriting files
547 from the beginning, thus creating a 'rotating' buffer.
548 In addition, it will name
549 the files with enough leading 0s to support the maximum number of
550 files, allowing them to sort correctly.
552 Used in conjunction with the
554 option, this will limit the number of rotated dump files that get
555 created, exiting with status 0 when reaching the limit. If used with
557 as well, the behavior will result in cyclical files per timeslice.
560 When parsing and printing,
561 in addition to printing the headers of each packet, print the data of
562 each packet (minus its link level header) in hex.
563 The smaller of the entire packet or
565 bytes will be printed. Note that this is the entire link-layer
566 packet, so for link layers that pad (e.g. Ethernet), the padding bytes
567 will also be printed when the higher layer packet is shorter than the
571 When parsing and printing,
572 in addition to printing the headers of each packet, print the data of
575 its link level header, in hex.
578 When parsing and printing,
579 in addition to printing the headers of each packet, print the data of
580 each packet (minus its link level header) in hex and ASCII.
581 This is very handy for analysing new protocols.
584 When parsing and printing,
585 in addition to printing the headers of each packet, print the data of
588 its link level header, in hex and ASCII.
591 Set the data link type to use while capturing packets to \fIdatalinktype\fP.
594 Used in conjunction with the
598 options, this will make
604 is the savefile being closed after each rotation. For example, specifying
608 will compress each savefile using gzip or bzip2.
610 Note that tcpdump will run the command in parallel to the capture, using
611 the lowest priority so that this doesn't disturb the capture process.
613 And in case you would like to use a command that itself takes flags or
614 different arguments, you can always write a shell script that will take the
615 savefile name as the only argument, make the flags & arguments arrangements
616 and execute the command that you want.
619 Drops privileges (if root) and changes user ID to
621 and the group ID to the primary group of
624 This behavior can also be enabled by default at compile time.
625 .IP "\fI expression\fP"
627 selects which packets will be dumped.
628 If no \fIexpression\fP
629 is given, all packets on the net will be dumped.
631 only packets for which \fIexpression\fP is `true' will be dumped.
633 For the \fIexpression\fP syntax, see
636 Expression arguments can be passed to \fItcpdump\fP as either a single
637 argument or as multiple arguments, whichever is more convenient.
638 Generally, if the expression contains Shell metacharacters, it is
639 easier to pass it as a single, quoted argument.
640 Multiple arguments are concatenated with spaces before being parsed.
643 To print all packets arriving at or departing from \fIsundown\fP:
646 \fBtcpdump host sundown\fP
650 To print traffic between \fIhelios\fR and either \fIhot\fR or \fIace\fR:
653 \fBtcpdump host helios and \\( hot or ace \\)\fP
657 To print all IP packets between \fIace\fR and any host except \fIhelios\fR:
660 \fBtcpdump ip host ace and not helios\fP
664 To print all traffic between local hosts and hosts at Berkeley:
668 tcpdump net ucb-ether
672 To print all ftp traffic through internet gateway \fIsnup\fP:
673 (note that the expression is quoted to prevent the shell from
674 (mis-)interpreting the parentheses):
678 tcpdump 'gateway snup and (port ftp or ftp-data)'
682 To print traffic neither sourced from nor destined for local hosts
683 (if you gateway to one other net, this stuff should never make it
684 onto your local net).
688 tcpdump ip and not net \fIlocalnet\fP
692 To print the start and end packets (the SYN and FIN packets) of each
693 TCP conversation that involves a non-local host.
697 tcpdump 'tcp[tcpflags] & (tcp-syn|tcp-fin) != 0 and not src and dst net \fIlocalnet\fP'
701 To print all IPv4 HTTP packets to and from port 80, i.e. print only
702 packets that contain data, not, for example, SYN and FIN packets and
703 ACK-only packets. (IPv6 is left as an exercise for the reader.)
707 tcpdump 'tcp port 80 and (((ip[2:2] - ((ip[0]&0xf)<<2)) - ((tcp[12]&0xf0)>>2)) != 0)'
711 To print IP packets longer than 576 bytes sent through gateway \fIsnup\fP:
715 tcpdump 'gateway snup and ip[2:2] > 576'
719 To print IP broadcast or multicast packets that were
721 sent via Ethernet broadcast or multicast:
725 tcpdump 'ether[0] & 1 = 0 and ip[16] >= 224'
729 To print all ICMP packets that are not echo requests/replies (i.e., not
734 tcpdump 'icmp[icmptype] != icmp-echo and icmp[icmptype] != icmp-echoreply'
739 The output of \fItcpdump\fP is protocol dependent.
741 gives a brief description and examples of most of the formats.
749 If the '-e' option is given, the link level header is printed out.
750 On Ethernets, the source and destination addresses, protocol,
751 and packet length are printed.
753 On FDDI networks, the '-e' option causes \fItcpdump\fP to print
754 the `frame control' field, the source and destination addresses,
755 and the packet length.
756 (The `frame control' field governs the
757 interpretation of the rest of the packet.
759 as those containing IP datagrams) are `async' packets, with a priority
760 value between 0 and 7; for example, `\fBasync4\fR'.
762 are assumed to contain an 802.2 Logical Link Control (LLC) packet;
763 the LLC header is printed if it is \fInot\fR an ISO datagram or a
764 so-called SNAP packet.
766 On Token Ring networks, the '-e' option causes \fItcpdump\fP to print
767 the `access control' and `frame control' fields, the source and
768 destination addresses, and the packet length.
770 packets are assumed to contain an LLC packet.
771 Regardless of whether
772 the '-e' option is specified or not, the source routing information is
773 printed for source-routed packets.
775 On 802.11 networks, the '-e' option causes \fItcpdump\fP to print
776 the `frame control' fields, all of the addresses in the 802.11 header,
777 and the packet length.
779 packets are assumed to contain an LLC packet.
781 \fI(N.B.: The following description assumes familiarity with
782 the SLIP compression algorithm described in RFC-1144.)\fP
784 On SLIP links, a direction indicator (``I'' for inbound, ``O'' for outbound),
785 packet type, and compression information are printed out.
786 The packet type is printed first.
787 The three types are \fIip\fP, \fIutcp\fP, and \fIctcp\fP.
788 No further link information is printed for \fIip\fR packets.
789 For TCP packets, the connection identifier is printed following the type.
790 If the packet is compressed, its encoded header is printed out.
791 The special cases are printed out as
792 \fB*S+\fIn\fR and \fB*SA+\fIn\fR, where \fIn\fR is the amount by which
793 the sequence number (or sequence number and ack) has changed.
794 If it is not a special case,
795 zero or more changes are printed.
796 A change is indicated by U (urgent pointer), W (window), A (ack),
797 S (sequence number), and I (packet ID), followed by a delta (+n or -n),
799 Finally, the amount of data in the packet and compressed header length
802 For example, the following line shows an outbound compressed TCP packet,
803 with an implicit connection identifier; the ack has changed by 6,
804 the sequence number by 49, and the packet ID by 6; there are 3 bytes of
805 data and 6 bytes of compressed header:
808 \fBO ctcp * A+6 S+49 I+6 3 (6)\fP
814 Arp/rarp output shows the type of request and its arguments.
816 format is intended to be self explanatory.
817 Here is a short sample taken from the start of an `rlogin' from
818 host \fIrtsg\fP to host \fIcsam\fP:
822 \f(CWarp who-has csam tell rtsg
823 arp reply csam is-at CSAM\fR
827 The first line says that rtsg sent an arp packet asking
828 for the Ethernet address of internet host csam.
830 replies with its Ethernet address (in this example, Ethernet addresses
831 are in caps and internet addresses in lower case).
833 This would look less redundant if we had done \fItcpdump \-n\fP:
837 \f(CWarp who-has 128.3.254.6 tell 128.3.254.68
838 arp reply 128.3.254.6 is-at 02:07:01:00:01:c4\fP
842 If we had done \fItcpdump \-e\fP, the fact that the first packet is
843 broadcast and the second is point-to-point would be visible:
847 \f(CWRTSG Broadcast 0806 64: arp who-has csam tell rtsg
848 CSAM RTSG 0806 64: arp reply csam is-at CSAM\fR
852 For the first packet this says the Ethernet source address is RTSG, the
853 destination is the Ethernet broadcast address, the type field
854 contained hex 0806 (type ETHER_ARP) and the total length was 64 bytes.
858 \fI(N.B.:The following description assumes familiarity with
859 the TCP protocol described in RFC-793.
860 If you are not familiar
861 with the protocol, neither this description nor \fItcpdump\fP will
862 be of much use to you.)\fP
864 The general format of a tcp protocol line is:
868 \fIsrc > dst: flags data-seqno ack window urgent options\fP
872 \fISrc\fP and \fIdst\fP are the source and destination IP
874 \fIFlags\fP are some combination of S (SYN),
875 F (FIN), P (PUSH), R (RST), W (ECN CWR) or E (ECN-Echo), or a single
877 \fIData-seqno\fP describes the portion of sequence space covered
878 by the data in this packet (see example below).
879 \fIAck\fP is sequence number of the next data expected the other
880 direction on this connection.
881 \fIWindow\fP is the number of bytes of receive buffer space available
882 the other direction on this connection.
883 \fIUrg\fP indicates there is `urgent' data in the packet.
884 \fIOptions\fP are tcp options enclosed in angle brackets (e.g., <mss 1024>).
886 \fISrc, dst\fP and \fIflags\fP are always present.
888 depend on the contents of the packet's tcp protocol header and
889 are output only if appropriate.
891 Here is the opening portion of an rlogin from host \fIrtsg\fP to
896 \s-2\f(CWrtsg.1023 > csam.login: S 768512:768512(0) win 4096 <mss 1024>
897 csam.login > rtsg.1023: S 947648:947648(0) ack 768513 win 4096 <mss 1024>
898 rtsg.1023 > csam.login: . ack 1 win 4096
899 rtsg.1023 > csam.login: P 1:2(1) ack 1 win 4096
900 csam.login > rtsg.1023: . ack 2 win 4096
901 rtsg.1023 > csam.login: P 2:21(19) ack 1 win 4096
902 csam.login > rtsg.1023: P 1:2(1) ack 21 win 4077
903 csam.login > rtsg.1023: P 2:3(1) ack 21 win 4077 urg 1
904 csam.login > rtsg.1023: P 3:4(1) ack 21 win 4077 urg 1\fR\s+2
908 The first line says that tcp port 1023 on rtsg sent a packet
911 The \fBS\fP indicates that the \fISYN\fP flag was set.
912 The packet sequence number was 768512 and it contained no data.
913 (The notation is `first:last(nbytes)' which means `sequence
915 up to but not including \fIlast\fP which is \fInbytes\fP bytes of user data'.)
916 There was no piggy-backed ack, the available receive window was 4096
917 bytes and there was a max-segment-size option requesting an mss of
920 Csam replies with a similar packet except it includes a piggy-backed
922 Rtsg then acks csam's SYN.
925 The packet contained no data so there is no data sequence number.
926 Note that the ack sequence
927 number is a small integer (1).
928 The first time \fItcpdump\fP sees a
929 tcp `conversation', it prints the sequence number from the packet.
930 On subsequent packets of the conversation, the difference between
931 the current packet's sequence number and this initial sequence number
933 This means that sequence numbers after the
934 first can be interpreted
935 as relative byte positions in the conversation's data stream (with the
936 first data byte each direction being `1').
937 `-S' will override this
938 feature, causing the original sequence numbers to be output.
940 On the 6th line, rtsg sends csam 19 bytes of data (bytes 2 through 20
941 in the rtsg \(-> csam side of the conversation).
942 The PUSH flag is set in the packet.
943 On the 7th line, csam says it's received data sent by rtsg up to
944 but not including byte 21.
945 Most of this data is apparently sitting in the
946 socket buffer since csam's receive window has gotten 19 bytes smaller.
947 Csam also sends one byte of data to rtsg in this packet.
948 On the 8th and 9th lines,
949 csam sends two bytes of urgent, pushed data to rtsg.
951 If the snapshot was small enough that \fItcpdump\fP didn't capture
952 the full TCP header, it interprets as much of the header as it can
953 and then reports ``[|\fItcp\fP]'' to indicate the remainder could not
955 If the header contains a bogus option (one with a length
956 that's either too small or beyond the end of the header), \fItcpdump\fP
957 reports it as ``[\fIbad opt\fP]'' and does not interpret any further
958 options (since it's impossible to tell where they start).
960 length indicates options are present but the IP datagram length is not
961 long enough for the options to actually be there, \fItcpdump\fP reports
962 it as ``[\fIbad hdr length\fP]''.
964 .B Capturing TCP packets with particular flag combinations (SYN-ACK, URG-ACK, etc.)
966 There are 8 bits in the control bits section of the TCP header:
968 .I CWR | ECE | URG | ACK | PSH | RST | SYN | FIN
970 Let's assume that we want to watch packets used in establishing
972 Recall that TCP uses a 3-way handshake protocol
973 when it initializes a new connection; the connection sequence with
974 regard to the TCP control bits is
980 2) Recipient responds with SYN, ACK
986 Now we're interested in capturing packets that have only the
987 SYN bit set (Step 1).
988 Note that we don't want packets from step 2
989 (SYN-ACK), just a plain initial SYN.
990 What we need is a correct filter
991 expression for \fItcpdump\fP.
993 Recall the structure of a TCP header without options:
997 -----------------------------------------------------------------
998 | source port | destination port |
999 -----------------------------------------------------------------
1001 -----------------------------------------------------------------
1002 | acknowledgment number |
1003 -----------------------------------------------------------------
1004 | HL | rsvd |C|E|U|A|P|R|S|F| window size |
1005 -----------------------------------------------------------------
1006 | TCP checksum | urgent pointer |
1007 -----------------------------------------------------------------
1010 A TCP header usually holds 20 octets of data, unless options are
1012 The first line of the graph contains octets 0 - 3, the
1013 second line shows octets 4 - 7 etc.
1015 Starting to count with 0, the relevant TCP control bits are contained
1020 ----------------|---------------|---------------|----------------
1021 | HL | rsvd |C|E|U|A|P|R|S|F| window size |
1022 ----------------|---------------|---------------|----------------
1023 | | 13th octet | | |
1026 Let's have a closer look at octet no. 13:
1036 These are the TCP control bits we are interested
1038 We have numbered the bits in this octet from 0 to 7, right to
1039 left, so the PSH bit is bit number 3, while the URG bit is number 5.
1041 Recall that we want to capture packets with only SYN set.
1042 Let's see what happens to octet 13 if a TCP datagram arrives
1043 with the SYN bit set in its header:
1054 control bits section we see that only bit number 1 (SYN) is set.
1056 Assuming that octet number 13 is an 8-bit unsigned integer in
1057 network byte order, the binary value of this octet is
1061 and its decimal representation is
1065 0*2 + 0*2 + 0*2 + 0*2 + 0*2 + 0*2 + 1*2 + 0*2 = 2
1068 We're almost done, because now we know that if only SYN is set,
1069 the value of the 13th octet in the TCP header, when interpreted
1070 as a 8-bit unsigned integer in network byte order, must be exactly 2.
1072 This relationship can be expressed as
1078 We can use this expression as the filter for \fItcpdump\fP in order
1079 to watch packets which have only SYN set:
1082 tcpdump -i xl0 tcp[13] == 2
1085 The expression says "let the 13th octet of a TCP datagram have
1086 the decimal value 2", which is exactly what we want.
1088 Now, let's assume that we need to capture SYN packets, but we
1089 don't care if ACK or any other TCP control bit is set at the
1091 Let's see what happens to octet 13 when a TCP datagram
1092 with SYN-ACK set arrives:
1102 Now bits 1 and 4 are set in the 13th octet.
1108 which translates to decimal
1112 0*2 + 0*2 + 0*2 + 1*2 + 0*2 + 0*2 + 1*2 + 0*2 = 18
1115 Now we can't just use 'tcp[13] == 18' in the \fItcpdump\fP filter
1116 expression, because that would select only those packets that have
1117 SYN-ACK set, but not those with only SYN set.
1118 Remember that we don't care
1119 if ACK or any other control bit is set as long as SYN is set.
1121 In order to achieve our goal, we need to logically AND the
1122 binary value of octet 13 with some other value to preserve
1124 We know that we want SYN to be set in any case,
1125 so we'll logically AND the value in the 13th octet with
1126 the binary value of a SYN:
1130 00010010 SYN-ACK 00000010 SYN
1131 AND 00000010 (we want SYN) AND 00000010 (we want SYN)
1133 = 00000010 = 00000010
1136 We see that this AND operation delivers the same result
1137 regardless whether ACK or another TCP control bit is set.
1138 The decimal representation of the AND value as well as
1139 the result of this operation is 2 (binary 00000010),
1140 so we know that for packets with SYN set the following
1141 relation must hold true:
1143 ( ( value of octet 13 ) AND ( 2 ) ) == ( 2 )
1145 This points us to the \fItcpdump\fP filter expression
1148 tcpdump -i xl0 'tcp[13] & 2 == 2'
1151 Some offsets and field values may be expressed as names
1152 rather than as numeric values. For example tcp[13] may
1153 be replaced with tcp[tcpflags]. The following TCP flag
1154 field values are also available: tcp-fin, tcp-syn, tcp-rst,
1155 tcp-push, tcp-act, tcp-urg.
1157 This can be demonstrated as:
1160 tcpdump -i xl0 'tcp[tcpflags] & tcp-push != 0'
1163 Note that you should use single quotes or a backslash
1164 in the expression to hide the AND ('&') special character
1170 UDP format is illustrated by this rwho packet:
1174 \f(CWactinide.who > broadcast.who: udp 84\fP
1178 This says that port \fIwho\fP on host \fIactinide\fP sent a udp
1179 datagram to port \fIwho\fP on host \fIbroadcast\fP, the Internet
1181 The packet contained 84 bytes of user data.
1183 Some UDP services are recognized (from the source or destination
1184 port number) and the higher level protocol information printed.
1185 In particular, Domain Name service requests (RFC-1034/1035) and Sun
1186 RPC calls (RFC-1050) to NFS.
1188 UDP Name Server Requests
1190 \fI(N.B.:The following description assumes familiarity with
1191 the Domain Service protocol described in RFC-1035.
1192 If you are not familiar
1193 with the protocol, the following description will appear to be written
1196 Name server requests are formatted as
1200 \fIsrc > dst: id op? flags qtype qclass name (len)\fP
1202 \f(CWh2opolo.1538 > helios.domain: 3+ A? ucbvax.berkeley.edu. (37)\fR
1206 Host \fIh2opolo\fP asked the domain server on \fIhelios\fP for an
1207 address record (qtype=A) associated with the name \fIucbvax.berkeley.edu.\fP
1208 The query id was `3'.
1209 The `+' indicates the \fIrecursion desired\fP flag
1211 The query length was 37 bytes, not including the UDP and
1212 IP protocol headers.
1213 The query operation was the normal one, \fIQuery\fP,
1214 so the op field was omitted.
1215 If the op had been anything else, it would
1216 have been printed between the `3' and the `+'.
1217 Similarly, the qclass was the normal one,
1218 \fIC_IN\fP, and omitted.
1219 Any other qclass would have been printed
1220 immediately after the `A'.
1222 A few anomalies are checked and may result in extra fields enclosed in
1223 square brackets: If a query contains an answer, authority records or
1224 additional records section,
1229 are printed as `[\fIn\fPa]', `[\fIn\fPn]' or `[\fIn\fPau]' where \fIn\fP
1230 is the appropriate count.
1231 If any of the response bits are set (AA, RA or rcode) or any of the
1232 `must be zero' bits are set in bytes two and three, `[b2&3=\fIx\fP]'
1233 is printed, where \fIx\fP is the hex value of header bytes two and three.
1235 UDP Name Server Responses
1237 Name server responses are formatted as
1241 \fIsrc > dst: id op rcode flags a/n/au type class data (len)\fP
1243 \f(CWhelios.domain > h2opolo.1538: 3 3/3/7 A 128.32.137.3 (273)
1244 helios.domain > h2opolo.1537: 2 NXDomain* 0/1/0 (97)\fR
1248 In the first example, \fIhelios\fP responds to query id 3 from \fIh2opolo\fP
1249 with 3 answer records, 3 name server records and 7 additional records.
1250 The first answer record is type A (address) and its data is internet
1251 address 128.32.137.3.
1252 The total size of the response was 273 bytes,
1253 excluding UDP and IP headers.
1254 The op (Query) and response code
1255 (NoError) were omitted, as was the class (C_IN) of the A record.
1257 In the second example, \fIhelios\fP responds to query 2 with a
1258 response code of non-existent domain (NXDomain) with no answers,
1259 one name server and no authority records.
1260 The `*' indicates that
1261 the \fIauthoritative answer\fP bit was set.
1263 answers, no type, class or data were printed.
1265 Other flag characters that might appear are `\-' (recursion available,
1266 RA, \fInot\fP set) and `|' (truncated message, TC, set).
1268 `question' section doesn't contain exactly one entry, `[\fIn\fPq]'
1274 \fItcpdump\fP now includes fairly extensive SMB/CIFS/NBT decoding for data
1275 on UDP/137, UDP/138 and TCP/139.
1276 Some primitive decoding of IPX and
1277 NetBEUI SMB data is also done.
1279 By default a fairly minimal decode is done, with a much more detailed
1280 decode done if -v is used.
1281 Be warned that with -v a single SMB packet
1282 may take up a page or more, so only use -v if you really want all the
1285 For information on SMB packet formats and what all te fields mean see
1286 www.cifs.org or the pub/samba/specs/ directory on your favorite
1287 samba.org mirror site.
1288 The SMB patches were written by Andrew Tridgell
1292 NFS Requests and Replies
1294 Sun NFS (Network File System) requests and replies are printed as:
1298 \fIsrc.xid > dst.nfs: len op args\fP
1299 \fIsrc.nfs > dst.xid: reply stat len op results\fP
1302 sushi.6709 > wrl.nfs: 112 readlink fh 21,24/10.73165
1303 wrl.nfs > sushi.6709: reply ok 40 readlink "../var"
1304 sushi.201b > wrl.nfs:
1305 144 lookup fh 9,74/4096.6878 "xcolors"
1306 wrl.nfs > sushi.201b:
1307 reply ok 128 lookup fh 9,74/4134.3150
1312 In the first line, host \fIsushi\fP sends a transaction with id \fI6709\fP
1313 to \fIwrl\fP (note that the number following the src host is a
1314 transaction id, \fInot\fP the source port).
1315 The request was 112 bytes,
1316 excluding the UDP and IP headers.
1317 The operation was a \fIreadlink\fP
1318 (read symbolic link) on file handle (\fIfh\fP) 21,24/10.731657119.
1319 (If one is lucky, as in this case, the file handle can be interpreted
1320 as a major,minor device number pair, followed by the inode number and
1322 \fIWrl\fP replies `ok' with the contents of the link.
1324 In the third line, \fIsushi\fP asks \fIwrl\fP to lookup the name
1325 `\fIxcolors\fP' in directory file 9,74/4096.6878.
1326 Note that the data printed
1327 depends on the operation type.
1328 The format is intended to be self
1329 explanatory if read in conjunction with
1330 an NFS protocol spec.
1332 If the \-v (verbose) flag is given, additional information is printed.
1338 sushi.1372a > wrl.nfs:
1339 148 read fh 21,11/12.195 8192 bytes @ 24576
1340 wrl.nfs > sushi.1372a:
1341 reply ok 1472 read REG 100664 ids 417/0 sz 29388
1346 (\-v also prints the IP header TTL, ID, length, and fragmentation fields,
1347 which have been omitted from this example.) In the first line,
1348 \fIsushi\fP asks \fIwrl\fP to read 8192 bytes from file 21,11/12.195,
1349 at byte offset 24576.
1350 \fIWrl\fP replies `ok'; the packet shown on the
1351 second line is the first fragment of the reply, and hence is only 1472
1352 bytes long (the other bytes will follow in subsequent fragments, but
1353 these fragments do not have NFS or even UDP headers and so might not be
1354 printed, depending on the filter expression used).
1355 Because the \-v flag
1356 is given, some of the file attributes (which are returned in addition
1357 to the file data) are printed: the file type (``REG'', for regular file),
1358 the file mode (in octal), the uid and gid, and the file size.
1360 If the \-v flag is given more than once, even more details are printed.
1362 Note that NFS requests are very large and much of the detail won't be printed
1363 unless \fIsnaplen\fP is increased.
1364 Try using `\fB\-s 192\fP' to watch
1367 NFS reply packets do not explicitly identify the RPC operation.
1369 \fItcpdump\fP keeps track of ``recent'' requests, and matches them to the
1370 replies using the transaction ID.
1371 If a reply does not closely follow the
1372 corresponding request, it might not be parsable.
1374 AFS Requests and Replies
1376 Transarc AFS (Andrew File System) requests and replies are printed
1382 \fIsrc.sport > dst.dport: rx packet-type\fP
1383 \fIsrc.sport > dst.dport: rx packet-type service call call-name args\fP
1384 \fIsrc.sport > dst.dport: rx packet-type service reply call-name args\fP
1387 elvis.7001 > pike.afsfs:
1388 rx data fs call rename old fid 536876964/1/1 ".newsrc.new"
1389 new fid 536876964/1/1 ".newsrc"
1390 pike.afsfs > elvis.7001: rx data fs reply rename
1395 In the first line, host elvis sends a RX packet to pike.
1397 a RX data packet to the fs (fileserver) service, and is the start of
1399 The RPC call was a rename, with the old directory file id
1400 of 536876964/1/1 and an old filename of `.newsrc.new', and a new directory
1401 file id of 536876964/1/1 and a new filename of `.newsrc'.
1403 responds with a RPC reply to the rename call (which was successful, because
1404 it was a data packet and not an abort packet).
1406 In general, all AFS RPCs are decoded at least by RPC call name.
1408 AFS RPCs have at least some of the arguments decoded (generally only
1409 the `interesting' arguments, for some definition of interesting).
1411 The format is intended to be self-describing, but it will probably
1412 not be useful to people who are not familiar with the workings of
1415 If the -v (verbose) flag is given twice, acknowledgement packets and
1416 additional header information is printed, such as the the RX call ID,
1417 call number, sequence number, serial number, and the RX packet flags.
1419 If the -v flag is given twice, additional information is printed,
1420 such as the the RX call ID, serial number, and the RX packet flags.
1421 The MTU negotiation information is also printed from RX ack packets.
1423 If the -v flag is given three times, the security index and service id
1426 Error codes are printed for abort packets, with the exception of Ubik
1427 beacon packets (because abort packets are used to signify a yes vote
1428 for the Ubik protocol).
1430 Note that AFS requests are very large and many of the arguments won't
1431 be printed unless \fIsnaplen\fP is increased.
1432 Try using `\fB-s 256\fP'
1433 to watch AFS traffic.
1435 AFS reply packets do not explicitly identify the RPC operation.
1437 \fItcpdump\fP keeps track of ``recent'' requests, and matches them to the
1438 replies using the call number and service ID.
1439 If a reply does not closely
1441 corresponding request, it might not be parsable.
1444 KIP AppleTalk (DDP in UDP)
1446 AppleTalk DDP packets encapsulated in UDP datagrams are de-encapsulated
1447 and dumped as DDP packets (i.e., all the UDP header information is
1451 is used to translate AppleTalk net and node numbers to names.
1452 Lines in this file have the form
1464 The first two lines give the names of AppleTalk networks.
1466 line gives the name of a particular host (a host is distinguished
1467 from a net by the 3rd octet in the number \-
1468 a net number \fImust\fP have two octets and a host number \fImust\fP
1469 have three octets.) The number and name should be separated by
1470 whitespace (blanks or tabs).
1473 file may contain blank lines or comment lines (lines starting with
1476 AppleTalk addresses are printed in the form
1482 \f(CW144.1.209.2 > icsd-net.112.220
1483 office.2 > icsd-net.112.220
1484 jssmag.149.235 > icsd-net.2\fR
1490 doesn't exist or doesn't contain an entry for some AppleTalk
1491 host/net number, addresses are printed in numeric form.)
1492 In the first example, NBP (DDP port 2) on net 144.1 node 209
1493 is sending to whatever is listening on port 220 of net icsd node 112.
1494 The second line is the same except the full name of the source node
1495 is known (`office').
1496 The third line is a send from port 235 on
1497 net jssmag node 149 to broadcast on the icsd-net NBP port (note that
1498 the broadcast address (255) is indicated by a net name with no host
1499 number \- for this reason it's a good idea to keep node names and
1500 net names distinct in /etc/atalk.names).
1502 NBP (name binding protocol) and ATP (AppleTalk transaction protocol)
1503 packets have their contents interpreted.
1504 Other protocols just dump
1505 the protocol name (or number if no name is registered for the
1506 protocol) and packet size.
1508 \fBNBP packets\fP are formatted like the following examples:
1512 \s-2\f(CWicsd-net.112.220 > jssmag.2: nbp-lkup 190: "=:LaserWriter@*"
1513 jssmag.209.2 > icsd-net.112.220: nbp-reply 190: "RM1140:LaserWriter@*" 250
1514 techpit.2 > icsd-net.112.220: nbp-reply 190: "techpit:LaserWriter@*" 186\fR\s+2
1518 The first line is a name lookup request for laserwriters sent by net icsd host
1519 112 and broadcast on net jssmag.
1520 The nbp id for the lookup is 190.
1521 The second line shows a reply for this request (note that it has the
1522 same id) from host jssmag.209 saying that it has a laserwriter
1523 resource named "RM1140" registered on port 250.
1525 another reply to the same request saying host techpit has laserwriter
1526 "techpit" registered on port 186.
1528 \fBATP packet\fP formatting is demonstrated by the following example:
1532 \s-2\f(CWjssmag.209.165 > helios.132: atp-req 12266<0-7> 0xae030001
1533 helios.132 > jssmag.209.165: atp-resp 12266:0 (512) 0xae040000
1534 helios.132 > jssmag.209.165: atp-resp 12266:1 (512) 0xae040000
1535 helios.132 > jssmag.209.165: atp-resp 12266:2 (512) 0xae040000
1536 helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
1537 helios.132 > jssmag.209.165: atp-resp 12266:4 (512) 0xae040000
1538 helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
1539 helios.132 > jssmag.209.165: atp-resp 12266:6 (512) 0xae040000
1540 helios.132 > jssmag.209.165: atp-resp*12266:7 (512) 0xae040000
1541 jssmag.209.165 > helios.132: atp-req 12266<3,5> 0xae030001
1542 helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
1543 helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
1544 jssmag.209.165 > helios.132: atp-rel 12266<0-7> 0xae030001
1545 jssmag.209.133 > helios.132: atp-req* 12267<0-7> 0xae030002\fR\s+2
1549 Jssmag.209 initiates transaction id 12266 with host helios by requesting
1550 up to 8 packets (the `<0-7>').
1551 The hex number at the end of the line
1552 is the value of the `userdata' field in the request.
1554 Helios responds with 8 512-byte packets.
1555 The `:digit' following the
1556 transaction id gives the packet sequence number in the transaction
1557 and the number in parens is the amount of data in the packet,
1558 excluding the atp header.
1559 The `*' on packet 7 indicates that the
1562 Jssmag.209 then requests that packets 3 & 5 be retransmitted.
1564 resends them then jssmag.209 releases the transaction.
1566 jssmag.209 initiates the next request.
1567 The `*' on the request
1568 indicates that XO (`exactly once') was \fInot\fP set.
1573 Fragmented Internet datagrams are printed as
1577 \fB(frag \fIid\fB:\fIsize\fB@\fIoffset\fB+)\fR
1578 \fB(frag \fIid\fB:\fIsize\fB@\fIoffset\fB)\fR
1582 (The first form indicates there are more fragments.
1584 indicates this is the last fragment.)
1586 \fIId\fP is the fragment id.
1587 \fISize\fP is the fragment
1588 size (in bytes) excluding the IP header.
1589 \fIOffset\fP is this
1590 fragment's offset (in bytes) in the original datagram.
1592 The fragment information is output for each fragment.
1594 fragment contains the higher level protocol header and the frag
1595 info is printed after the protocol info.
1597 after the first contain no higher level protocol header and the
1598 frag info is printed after the source and destination addresses.
1599 For example, here is part of an ftp from arizona.edu to lbl-rtsg.arpa
1600 over a CSNET connection that doesn't appear to handle 576 byte datagrams:
1604 \s-2\f(CWarizona.ftp-data > rtsg.1170: . 1024:1332(308) ack 1 win 4096 (frag 595a:328@0+)
1605 arizona > rtsg: (frag 595a:204@328)
1606 rtsg.1170 > arizona.ftp-data: . ack 1536 win 2560\fP\s+2
1610 There are a couple of things to note here: First, addresses in the
1611 2nd line don't include port numbers.
1612 This is because the TCP
1613 protocol information is all in the first fragment and we have no idea
1614 what the port or sequence numbers are when we print the later fragments.
1615 Second, the tcp sequence information in the first line is printed as if there
1616 were 308 bytes of user data when, in fact, there are 512 bytes (308 in
1617 the first frag and 204 in the second).
1618 If you are looking for holes
1619 in the sequence space or trying to match up acks
1620 with packets, this can fool you.
1622 A packet with the IP \fIdon't fragment\fP flag is marked with a
1623 trailing \fB(DF)\fP.
1627 By default, all output lines are preceded by a timestamp.
1629 is the current clock time in the form
1635 and is as accurate as the kernel's clock.
1636 The timestamp reflects the time the kernel first saw the packet.
1638 is made to account for the time lag between when the
1639 Ethernet interface removed the packet from the wire and when the kernel
1640 serviced the `new packet' interrupt.
1642 stty(1), pcap(3PCAP), bpf(4), nit(4P), pcap-savefile(5),
1645 The original authors are:
1649 Steven McCanne, all of the
1650 Lawrence Berkeley National Laboratory, University of California, Berkeley, CA.
1652 It is currently being maintained by tcpdump.org.
1654 The current version is available via http:
1657 .I http://www.tcpdump.org/
1660 The original distribution is available via anonymous ftp:
1663 .I ftp://ftp.ee.lbl.gov/tcpdump.tar.Z
1666 IPv6/IPsec support is added by WIDE/KAME project.
1667 This program uses Eric Young's SSLeay library, under specific configurations.
1669 Please send problems, bugs, questions, desirable enhancements, patches
1673 tcpdump-workers@lists.tcpdump.org
1676 NIT doesn't let you watch your own outbound traffic, BPF will.
1677 We recommend that you use the latter.
1679 On Linux systems with 2.0[.x] kernels:
1681 packets on the loopback device will be seen twice;
1683 packet filtering cannot be done in the kernel, so that all packets must
1684 be copied from the kernel in order to be filtered in user mode;
1686 all of a packet, not just the part that's within the snapshot length,
1687 will be copied from the kernel (the 2.0[.x] packet capture mechanism, if
1688 asked to copy only part of a packet to userland, will not report the
1689 true length of the packet; this would cause most IP packets to get an
1693 capturing on some PPP devices won't work correctly.
1695 We recommend that you upgrade to a 2.2 or later kernel.
1697 Some attempt should be made to reassemble IP fragments or, at least
1698 to compute the right length for the higher level protocol.
1700 Name server inverse queries are not dumped correctly: the (empty)
1701 question section is printed rather than real query in the answer
1703 Some believe that inverse queries are themselves a bug and
1704 prefer to fix the program generating them rather than \fItcpdump\fP.
1706 A packet trace that crosses a daylight savings time change will give
1707 skewed time stamps (the time change is ignored).
1709 Filter expressions on fields other than those in Token Ring headers will
1710 not correctly handle source-routed Token Ring packets.
1712 Filter expressions on fields other than those in 802.11 headers will not
1713 correctly handle 802.11 data packets with both To DS and From DS set.
1716 should chase header chain, but at this moment it does not.
1717 .BR "ip6 protochain"
1718 is supplied for this behavior.
1720 Arithmetic expression against transport layer headers, like \fBtcp[0]\fP,
1721 does not work against IPv6 packets.
1722 It only looks at IPv4 packets.
projects / FreeBSD / releng / 9.0.git / blob