1TCPDUMP(8) System Manager's Manual TCPDUMP(8)
2
6 tcpdump - dump traffic on a network
7
9 tcpdump [ -AbdDefhHIJKlLnNOpqStuUvxX# ] [ -B buffer_size ]
10 [ -c count ] [ --count ] [ -C file_size ]
11 [ -E spi@ipaddr algo:secret,... ]
12 [ -F file ] [ -G rotate_seconds ] [ -i interface ]
13 [ --immediate-mode ] [ -j tstamp_type ] [ -m module ]
14 [ -M secret ] [ --number ] [ --print ] [ -Q in|out|inout ]
15 [ -r file ] [ -s snaplen ] [ -T type ] [ --version ]
16 [ -V file ] [ -w file ] [ -W filecount ] [ -y datalinktype ]
17 [ -z postrotate-command ] [ -Z user ]
18 [ --time-stamp-precision=tstamp_precision ]
19 [ --micro ] [ --nano ]
20 [ expression ]
21
23 Tcpdump prints out a description of the contents of packets on a net‐
24 work interface that match the Boolean expression (see pcap-filter(7)
25 for the expression syntax); the description is preceded by a time
26 stamp, printed, by default, as hours, minutes, seconds, and fractions
27 of a second since midnight. It can also be run with the -w flag, which
28 causes it to save the packet data to a file for later analysis, and/or
29 with the -r flag, which causes it to read from a saved packet file
30 rather than to read packets from a network interface. It can also be
31 run with the -V flag, which causes it to read a list of saved packet
32 files. In all cases, only packets that match expression will be pro‐
33 cessed by tcpdump.
34 Tcpdump will, if not run with the -c flag, continue capturing packets
36 until it is interrupted by a SIGINT signal (generated, for example, by
37 typing your interrupt character, typically control-C) or a SIGTERM sig‐
38 nal (typically generated with the kill(1) command); if run with the -c
39 flag, it will capture packets until it is interrupted by a SIGINT or
40 SIGTERM signal or the specified number of packets have been processed.
41 When tcpdump finishes capturing packets, it will report counts of:
43 packets ``captured'' (this is the number of packets that tcpdump
45 has received and processed);
46 packets ``received by filter'' (the meaning of this depends on
48 the OS on which you're running tcpdump, and possibly on the way
49 the OS was configured - if a filter was specified on the command
50 line, on some OSes it counts packets regardless of whether they
51 were matched by the filter expression and, even if they were
52 matched by the filter expression, regardless of whether tcpdump
53 has read and processed them yet, on other OSes it counts only
54 packets that were matched by the filter expression regardless of
55 whether tcpdump has read and processed them yet, and on other
56 OSes it counts only packets that were matched by the filter ex‐
57 pression and were processed by tcpdump);
58 packets ``dropped by kernel'' (this is the number of packets
60 that were dropped, due to a lack of buffer space, by the packet
61 capture mechanism in the OS on which tcpdump is running, if the
62 OS reports that information to applications; if not, it will be
63 reported as 0).
64 On platforms that support the SIGINFO signal, such as most BSDs (in‐
66 cluding macOS) and Digital/Tru64 UNIX, it will report those counts when
67 it receives a SIGINFO signal (generated, for example, by typing your
68 ``status'' character, typically control-T, although on some platforms,
69 such as macOS, the ``status'' character is not set by default, so you
70 must set it with stty(1) in order to use it) and will continue captur‐
71 ing packets. On platforms that do not support the SIGINFO signal, the
72 same can be achieved by using the SIGUSR1 signal.
73 Using the SIGUSR2 signal along with the -w flag will forcibly flush the
75 packet buffer into the output file.
76 Reading packets from a network interface may require that you have spe‐
78 cial privileges; see the pcap(3PCAP) man page for details. Reading a
79 saved packet file doesn't require special privileges.
80
82 -A Print each packet (minus its link level header) in ASCII. Handy
83 for capturing web pages.
84 -b Print the AS number in BGP packets in ASDOT notation rather than
86 ASPLAIN notation.
87 -B buffer_size
89 --buffer-size=buffer_size
90 Set the operating system capture buffer size to buffer_size, in
91 units of KiB (1024 bytes).
92 -c count
94 Exit after receiving count packets.
95 --count
97 Print only on stdout the packet count when reading capture
98 file(s) instead of parsing/printing the packets. If a filter is
99 specified on the command line, tcpdump counts only packets that
100 were matched by the filter expression.
101 -C file_size
103 Before writing a raw packet to a savefile, check whether the
104 file is currently larger than file_size and, if so, close the
105 current savefile and open a new one. Savefiles after the first
106 savefile will have the name specified with the -w flag, with a
107 number after it, starting at 1 and continuing upward. The units
108 of file_size are millions of bytes (1,000,000 bytes, not
109 1,048,576 bytes).
110 Note that when used with -Z option (enabled by default), privi‐
112 leges are dropped before opening the first savefile.
113 -d Dump the compiled packet-matching code in a human readable form
115 to standard output and stop.
116 Please mind that although code compilation is always DLT-spe‐
118 cific, typically it is impossible (and unnecessary) to specify
119 which DLT to use for the dump because tcpdump uses either the
120 DLT of the input pcap file specified with -r, or the default DLT
121 of the network interface specified with -i, or the particular
122 DLT of the network interface specified with -y and -i respec‐
123 tively. In these cases the dump shows the same exact code that
124 would filter the input file or the network interface without -d.
125 However, when neither -r nor -i is specified, specifying -d pre‐
127 vents tcpdump from guessing a suitable network interface (see
128 -i). In this case the DLT defaults to EN10MB and can be set to
129 another valid value manually with -y.
130 -dd Dump packet-matching code as a C program fragment.
132 -ddd Dump packet-matching code as decimal numbers (preceded with a
134 count).
135 -D
137 --list-interfaces
138 Print the list of the network interfaces available on the system
139 and on which tcpdump can capture packets. For each network in‐
140 terface, a number and an interface name, possibly followed by a
141 text description of the interface, are printed. The interface
142 name or the number can be supplied to the -i flag to specify an
143 interface on which to capture.
144 This can be useful on systems that don't have a command to list
146 them (e.g., Windows systems, or UNIX systems lacking ifconfig
147 -a); the number can be useful on Windows 2000 and later systems,
148 where the interface name is a somewhat complex string.
149 The -D flag will not be supported if tcpdump was built with an
151 older version of libpcap that lacks the pcap_findalldevs(3PCAP)
152 function.
153 -e Print the link-level header on each dump line. This can be
155 used, for example, to print MAC layer addresses for protocols
156 such as Ethernet and IEEE 802.11.
157 -E Use spi@ipaddr algo:secret for decrypting IPsec ESP packets that
159 are addressed to addr and contain Security Parameter Index value
160 spi. This combination may be repeated with comma or newline sep‐
161 aration.
162 Note that setting the secret for IPv4 ESP packets is supported
164 at this time.
165 Algorithms may be des-cbc, 3des-cbc, blowfish-cbc, rc3-cbc,
167 cast128-cbc, or none. The default is des-cbc. The ability to
168 decrypt packets is only present if tcpdump was compiled with
169 cryptography enabled.
170 secret is the ASCII text for ESP secret key. If preceded by 0x,
172 then a hex value will be read.
173 The option assumes RFC 2406 ESP, not RFC 1827 ESP. The option
175 is only for debugging purposes, and the use of this option with
176 a true `secret' key is discouraged. By presenting IPsec secret
177 key onto command line you make it visible to others, via ps(1)
178 and other occasions.
179 In addition to the above syntax, the syntax file name may be
181 used to have tcpdump read the provided file in. The file is
182 opened upon receiving the first ESP packet, so any special per‐
183 missions that tcpdump may have been given should already have
184 been given up.
185 -f Print `foreign' IPv4 addresses numerically rather than symboli‐
187 cally (this option is intended to get around serious brain dam‐
188 age in Sun's NIS server — usually it hangs forever translating
189 non-local internet numbers).
190 The test for `foreign' IPv4 addresses is done using the IPv4 ad‐
192 dress and netmask of the interface on that capture is being
193 done. If that address or netmask are not available, either be‐
194 cause the interface on that capture is being done has no address
195 or netmask or because it is the "any" pseudo-interface, which is
196 available in Linux and in recent versions of macOS and Solaris,
197 and which can capture on more than one interface, this option
198 will not work correctly.
199 -F file
201 Use file as input for the filter expression. An additional ex‐
202 pression given on the command line is ignored.
203 -G rotate_seconds
205 If specified, rotates the dump file specified with the -w option
206 every rotate_seconds seconds. Savefiles will have the name
207 specified by -w which should include a time format as defined by
208 strftime(3). If no time format is specified, each new file will
209 overwrite the previous. Whenever a generated filename is not
210 unique, tcpdump will overwrite the pre-existing data; providing
211 a time specification that is coarser than the capture period is
212 therefore not advised.
213 If used in conjunction with the -C option, filenames will take
215 the form of `file<count>'.
216 Note that when used with -Z option (enabled by default), privi‐
218 leges are dropped before opening the first savefile.
219 -h
221 --help Print the tcpdump and libpcap version strings, print a usage
222 message, and exit.
223 --version
225 Print the tcpdump and libpcap version strings and exit.
226 -H Attempt to detect 802.11s draft mesh headers.
228 -i interface
230 --interface=interface
231 Listen, report the list of link-layer types, report the list of
232 time stamp types, or report the results of compiling a filter
233 expression on interface. If unspecified and if the -d flag is
234 not given, tcpdump searches the system interface list for the
235 lowest numbered, configured up interface (excluding loopback),
236 which may turn out to be, for example, ``eth0''.
237 On Linux systems with 2.2 or later kernels and on recent ver‐
239 sions of macOS and Solaris, an interface argument of ``any'' can
240 be used to capture packets from all interfaces. Note that cap‐
241 tures on the ``any'' pseudo-interface will not be done in pro‐
242 miscuous mode.
243 If the -D flag is supported, an interface number as printed by
245 that flag can be used as the interface argument, if no interface
246 on the system has that number as a name.
247 -I
249 --monitor-mode
250 Put the interface in "monitor mode"; this is supported only on
251 IEEE 802.11 Wi-Fi interfaces, and supported only on some operat‐
252 ing systems.
253 Note that in monitor mode the adapter might disassociate from
255 the network with which it's associated, so that you will not be
256 able to use any wireless networks with that adapter. This could
257 prevent accessing files on a network server, or resolving host
258 names or network addresses, if you are capturing in monitor mode
259 and are not connected to another network with another adapter.
260 This flag will affect the output of the -L flag. If -I isn't
262 specified, only those link-layer types available when not in
263 monitor mode will be shown; if -I is specified, only those link-
264 layer types available when in monitor mode will be shown.
265 --immediate-mode
267 Capture in "immediate mode". In this mode, packets are deliv‐
268 ered to tcpdump as soon as they arrive, rather than being
269 buffered for efficiency. This is the default when printing
270 packets rather than saving packets to a ``savefile'' if the
271 packets are being printed to a terminal rather than to a file or
272 pipe.
273 -j tstamp_type
275 --time-stamp-type=tstamp_type
276 Set the time stamp type for the capture to tstamp_type. The
277 names to use for the time stamp types are given in
278 pcap-tstamp(7); not all the types listed there will necessarily
279 be valid for any given interface.
280 -J
282 --list-time-stamp-types
283 List the supported time stamp types for the interface and exit.
284 If the time stamp type cannot be set for the interface, no time
285 stamp types are listed.
286 --time-stamp-precision=tstamp_precision
288 When capturing, set the time stamp precision for the capture to
289 tstamp_precision. Note that availability of high precision time
290 stamps (nanoseconds) and their actual accuracy is platform and
291 hardware dependent. Also note that when writing captures made
292 with nanosecond accuracy to a savefile, the time stamps are
293 written with nanosecond resolution, and the file is written with
294 a different magic number, to indicate that the time stamps are
295 in seconds and nanoseconds; not all programs that read pcap
296 savefiles will be able to read those captures.
297 When reading a savefile, convert time stamps to the precision
299 specified by timestamp_precision, and display them with that
300 resolution. If the precision specified is less than the preci‐
301 sion of time stamps in the file, the conversion will lose preci‐
302 sion.
303 The supported values for timestamp_precision are micro for mi‐
305 crosecond resolution and nano for nanosecond resolution. The
306 default is microsecond resolution.
307 --micro
309 --nano Shorthands for --time-stamp-precision=micro or --time-stamp-pre‐
310 cision=nano, adjusting the time stamp precision accordingly.
311 When reading packets from a savefile, using --micro truncates
312 time stamps if the savefile was created with nanosecond preci‐
313 sion. In contrast, a savefile created with microsecond preci‐
314 sion will have trailing zeroes added to the time stamp when
315 --nano is used.
316 -K
318 --dont-verify-checksums
319 Don't attempt to verify IP, TCP, or UDP checksums. This is use‐
320 ful for interfaces that perform some or all of those checksum
321 calculation in hardware; otherwise, all outgoing TCP checksums
322 will be flagged as bad.
323 -l Make stdout line buffered. Useful if you want to see the data
325 while capturing it. E.g.,
326 tcpdump -l | tee dat
328 or
330 tcpdump -l > dat & tail -f dat
332 Note that on Windows,``line buffered'' means ``unbuffered'', so
334 that WinDump will write each character individually if -l is
335 specified.
336 -U is similar to -l in its behavior, but it will cause output to
338 be ``packet-buffered'', so that the output is written to stdout
339 at the end of each packet rather than at the end of each line;
340 this is buffered on all platforms, including Windows.
341 -L
343 --list-data-link-types
344 List the known data link types for the interface, in the speci‐
345 fied mode, and exit. The list of known data link types may be
346 dependent on the specified mode; for example, on some platforms,
347 a Wi-Fi interface might support one set of data link types when
348 not in monitor mode (for example, it might support only fake
349 Ethernet headers, or might support 802.11 headers but not sup‐
350 port 802.11 headers with radio information) and another set of
351 data link types when in monitor mode (for example, it might sup‐
352 port 802.11 headers, or 802.11 headers with radio information,
353 only in monitor mode).
354 -m module
356 Load SMI MIB module definitions from file module. This option
357 can be used several times to load several MIB modules into tcp‐
358 dump.
359 -M secret
361 Use secret as a shared secret for validating the digests found
362 in TCP segments with the TCP-MD5 option (RFC 2385), if present.
363 -n Don't convert host addresses to names. This can be used to
365 avoid DNS lookups.
366 -nn Don't convert protocol and port numbers etc. to names either.
368 -N Don't print domain name qualification of host names. E.g., if
370 you give this flag then tcpdump will print ``nic'' instead of
371 ``nic.ddn.mil''.
372 -#
374 --number
375 Print an optional packet number at the beginning of the line.
376 -O
378 --no-optimize
379 Do not run the packet-matching code optimizer. This is useful
380 only if you suspect a bug in the optimizer.
381 -p
383 --no-promiscuous-mode
384 Don't put the interface into promiscuous mode. Note that the
385 interface might be in promiscuous mode for some other reason;
386 hence, `-p' cannot be used as an abbreviation for `ether host
387 {local-hw-addr} or ether broadcast'.
388 --print
390 Print parsed packet output, even if the raw packets are being
391 saved to a file with the -w flag.
392 -Q direction
394 --direction=direction
395 Choose send/receive direction direction for which packets should
396 be captured. Possible values are `in', `out' and `inout'. Not
397 available on all platforms.
398 -q Quick (quiet?) output. Print less protocol information so out‐
400 put lines are shorter.
401 -r file
403 Read packets from file (which was created with the -w option or
404 by other tools that write pcap or pcapng files). Standard input
405 is used if file is ``-''.
406 -S
408 --absolute-tcp-sequence-numbers
409 Print absolute, rather than relative, TCP sequence numbers.
410 -s snaplen
412 --snapshot-length=snaplen
413 Snarf snaplen bytes of data from each packet rather than the de‐
414 fault of 262144 bytes. Packets truncated because of a limited
415 snapshot are indicated in the output with ``[|proto]'', where
416 proto is the name of the protocol level at which the truncation
417 has occurred.
418 Note that taking larger snapshots both increases the amount of
420 time it takes to process packets and, effectively, decreases the
421 amount of packet buffering. This may cause packets to be lost.
422 Note also that taking smaller snapshots will discard data from
423 protocols above the transport layer, which loses information
424 that may be important. NFS and AFS requests and replies, for
425 example, are very large, and much of the detail won't be avail‐
426 able if a too-short snapshot length is selected.
427 If you need to reduce the snapshot size below the default, you
429 should limit snaplen to the smallest number that will capture
430 the protocol information you're interested in. Setting snaplen
431 to 0 sets it to the default of 262144, for backwards compatibil‐
432 ity with recent older versions of tcpdump.
433 -T type
435 Force packets selected by "expression" to be interpreted the
436 specified type. Currently known types are aodv (Ad-hoc On-de‐
437 mand Distance Vector protocol), carp (Common Address Redundancy
438 Protocol), cnfp (Cisco NetFlow protocol), domain (Domain Name
439 System), lmp (Link Management Protocol), pgm (Pragmatic General
440 Multicast), pgm_zmtp1 (ZMTP/1.0 inside PGM/EPGM), ptp (Precision
441 Time Protocol), radius (RADIUS), resp (REdis Serialization Pro‐
442 tocol), rpc (Remote Procedure Call), rtcp (Real-Time Applica‐
443 tions control protocol), rtp (Real-Time Applications protocol),
444 snmp (Simple Network Management Protocol), someip (SOME/IP),
445 tftp (Trivial File Transfer Protocol), vat (Visual Audio Tool),
446 vxlan (Virtual eXtensible Local Area Network), wb (distributed
447 White Board) and zmtp1 (ZeroMQ Message Transport Protocol 1.0).
448 Note that the pgm type above affects UDP interpretation only,
450 the native PGM is always recognised as IP protocol 113 regard‐
451 less. UDP-encapsulated PGM is often called "EPGM" or "PGM/UDP".
452 Note that the pgm_zmtp1 type above affects interpretation of
454 both native PGM and UDP at once. During the native PGM decoding
455 the application data of an ODATA/RDATA packet would be decoded
456 as a ZeroMQ datagram with ZMTP/1.0 frames. During the UDP de‐
457 coding in addition to that any UDP packet would be treated as an
458 encapsulated PGM packet.
459 -t Don't print a timestamp on each dump line.
461 -tt Print the timestamp, as seconds since January 1, 1970, 00:00:00,
463 UTC, and fractions of a second since that time, on each dump
464 line.
465 -ttt Print a delta (microsecond or nanosecond resolution depending on
467 the --time-stamp-precision option) between current and previous
468 line on each dump line. The default is microsecond resolution.
469 -tttt Print a timestamp, as hours, minutes, seconds, and fractions of
471 a second since midnight, preceded by the date, on each dump
472 line.
473 -ttttt Print a delta (microsecond or nanosecond resolution depending on
475 the --time-stamp-precision option) between current and first
476 line on each dump line. The default is microsecond resolution.
477 -u Print undecoded NFS handles.
479 -U
481 --packet-buffered
482 If the -w option is not specified, or if it is specified but the
483 --print flag is also specified, make the printed packet output
484 ``packet-buffered''; i.e., as the description of the contents of
485 each packet is printed, it will be written to the standard out‐
486 put, rather than, when not writing to a terminal, being written
487 only when the output buffer fills.
488 If the -w option is specified, make the saved raw packet output
490 ``packet-buffered''; i.e., as each packet is saved, it will be
491 written to the output file, rather than being written only when
492 the output buffer fills.
493 The -U flag will not be supported if tcpdump was built with an
495 older version of libpcap that lacks the pcap_dump_flush(3PCAP)
496 function.
497 -v When parsing and printing, produce (slightly more) verbose out‐
499 put. For example, the time to live, identification, total
500 length and options in an IP packet are printed. Also enables
501 additional packet integrity checks such as verifying the IP and
502 ICMP header checksum.
503 When writing to a file with the -w option and at the same time
505 not reading from a file with the -r option, report to stderr,
506 once per second, the number of packets captured. In Solaris,
507 FreeBSD and possibly other operating systems this periodic up‐
508 date currently can cause loss of captured packets on their way
509 from the kernel to tcpdump.
510 -vv Even more verbose output. For example, additional fields are
512 printed from NFS reply packets, and SMB packets are fully de‐
513 coded.
514 -vvv Even more verbose output. For example, telnet SB ... SE options
516 are printed in full. With -X Telnet options are printed in hex
517 as well.
518 -V file
520 Read a list of filenames from file. Standard input is used if
521 file is ``-''.
522 -w file
524 Write the raw packets to file rather than parsing and printing
525 them out. They can later be printed with the -r option. Stan‐
526 dard output is used if file is ``-''.
527 This output will be buffered if written to a file or pipe, so a
529 program reading from the file or pipe may not see packets for an
530 arbitrary amount of time after they are received. Use the -U
531 flag to cause packets to be written as soon as they are re‐
532 ceived.
533 The MIME type application/vnd.tcpdump.pcap has been registered
535 with IANA for pcap files. The filename extension .pcap appears
536 to be the most commonly used along with .cap and .dmp. Tcpdump
537 itself doesn't check the extension when reading capture files
538 and doesn't add an extension when writing them (it uses magic
539 numbers in the file header instead). However, many operating
540 systems and applications will use the extension if it is present
541 and adding one (e.g. .pcap) is recommended.
542 See pcap-savefile(5) for a description of the file format.
544 -W filecount
546 Used in conjunction with the -C option, this will limit the num‐
547 ber of files created to the specified number, and begin over‐
548 writing files from the beginning, thus creating a 'rotating'
549 buffer. In addition, it will name the files with enough leading
550 0s to support the maximum number of files, allowing them to sort
551 correctly.
552 Used in conjunction with the -G option, this will limit the num‐
554 ber of rotated dump files that get created, exiting with status
555 0 when reaching the limit.
556 If used in conjunction with both -C and -G, the -W option will
558 currently be ignored, and will only affect the file name.
559 -x When parsing and printing, in addition to printing the headers
561 of each packet, print the data of each packet (minus its link
562 level header) in hex. The smaller of the entire packet or
563 snaplen bytes will be printed. Note that this is the entire
564 link-layer packet, so for link layers that pad (e.g. Ethernet),
565 the padding bytes will also be printed when the higher layer
566 packet is shorter than the required padding. In the current im‐
567 plementation this flag may have the same effect as -xx if the
568 packet is truncated.
569 -xx When parsing and printing, in addition to printing the headers
571 of each packet, print the data of each packet, including its
572 link level header, in hex.
573 -X When parsing and printing, in addition to printing the headers
575 of each packet, print the data of each packet (minus its link
576 level header) in hex and ASCII. This is very handy for
577 analysing new protocols. In the current implementation this
578 flag may have the same effect as -XX if the packet is truncated.
579 -XX When parsing and printing, in addition to printing the headers
581 of each packet, print the data of each packet, including its
582 link level header, in hex and ASCII.
583 -y datalinktype
585 --linktype=datalinktype
586 Set the data link type to use while capturing packets (see -L)
587 or just compiling and dumping packet-matching code (see -d) to
588 datalinktype.
589 -z postrotate-command
591 Used in conjunction with the -C or -G options, this will make
592 tcpdump run " postrotate-command file " where file is the save‐
593 file being closed after each rotation. For example, specifying
594 -z gzip or -z bzip2 will compress each savefile using gzip or
595 bzip2.
596 Note that tcpdump will run the command in parallel to the cap‐
598 ture, using the lowest priority so that this doesn't disturb the
599 capture process.
600 And in case you would like to use a command that itself takes
602 flags or different arguments, you can always write a shell
603 script that will take the savefile name as the only argument,
604 make the flags & arguments arrangements and execute the command
605 that you want.
606 -Z user
608 --relinquish-privileges=user
609 If tcpdump is running as root, after opening the capture device
610 or input savefile, but before opening any savefiles for output,
611 change the user ID to user and the group ID to the primary group
612 of user.
613 This behavior is enabled by default (-Z tcpdump), and can be
615 disabled by -Z root.
616 expression
619 selects which packets will be dumped. If no expression is
620 given, all packets on the net will be dumped. Otherwise, only
621 packets for which expression is `true' will be dumped.
622 For the expression syntax, see pcap-filter(7).
624 The expression argument can be passed to tcpdump as either a
626 single Shell argument, or as multiple Shell arguments, whichever
627 is more convenient. Generally, if the expression contains Shell
628 metacharacters, such as backslashes used to escape protocol
629 names, it is easier to pass it as a single, quoted argument
630 rather than to escape the Shell metacharacters. Multiple argu‐
631 ments are concatenated with spaces before being parsed.
632
634 To print all packets arriving at or departing from sundown:
635 tcpdump host sundown
636 To print traffic between helios and either hot or ace:
638 tcpdump host helios and \( hot or ace \)
639 To print all IP packets between ace and any host except helios:
641 tcpdump ip host ace and not helios
642 To print all traffic between local hosts and hosts at Berkeley:
644 tcpdump net ucb-ether
645 To print all ftp traffic through internet gateway snup: (note that the
647 expression is quoted to prevent the shell from (mis-)interpreting the
648 parentheses):
649 tcpdump 'gateway snup and (port ftp or ftp-data)'
650 To print traffic neither sourced from nor destined for local hosts (if
652 you gateway to one other net, this stuff should never make it onto your
653 local net).
654 tcpdump ip and not net localnet
655 To print the start and end packets (the SYN and FIN packets) of each
657 TCP conversation that involves a non-local host.
658 tcpdump 'tcp[tcpflags] & (tcp-syn|tcp-fin) != 0 and not src and dst net localnet'
659 To print the TCP packets with flags RST and ACK both set. (i.e. select
661 only the RST and ACK flags in the flags field, and if the result is
662 "RST and ACK both set", match)
663 tcpdump 'tcp[tcpflags] & (tcp-rst|tcp-ack) == (tcp-rst|tcp-ack)'
664 To print all IPv4 HTTP packets to and from port 80, i.e. print only
666 packets that contain data, not, for example, SYN and FIN packets and
667 ACK-only packets. (IPv6 is left as an exercise for the reader.)
668 tcpdump 'tcp port 80 and (((ip[2:2] - ((ip[0]&0xf)<<2)) - ((tcp[12]&0xf0)>>2)) != 0)'
669 To print IP packets longer than 576 bytes sent through gateway snup:
671 tcpdump 'gateway snup and ip[2:2] > 576'
672 To print IP broadcast or multicast packets that were not sent via Eth‐
674 ernet broadcast or multicast:
675 tcpdump 'ether[0] & 1 = 0 and ip[16] >= 224'
676 To print all ICMP packets that are not echo requests/replies (i.e., not
678 ping packets):
679 tcpdump 'icmp[icmptype] != icmp-echo and icmp[icmptype] != icmp-echoreply'
680
682 The output of tcpdump is protocol dependent. The following gives a
683 brief description and examples of most of the formats.
684 Timestamps
686 By default, all output lines are preceded by a timestamp. The time‐
688 stamp is the current clock time in the form
689 hh:mm:ss.frac
690 and is as accurate as the kernel's clock. The timestamp reflects the
691 time the kernel applied a time stamp to the packet. No attempt is made
692 to account for the time lag between when the network interface finished
693 receiving the packet from the network and when the kernel applied a
694 time stamp to the packet; that time lag could include a delay between
695 the time when the network interface finished receiving a packet from
696 the network and the time when an interrupt was delivered to the kernel
697 to get it to read the packet and a delay between the time when the ker‐
698 nel serviced the `new packet' interrupt and the time when it applied a
699 time stamp to the packet.
700 Link Level Headers
702 If the '-e' option is given, the link level header is printed out. On
704 Ethernets, the source and destination addresses, protocol, and packet
705 length are printed.
706 On FDDI networks, the '-e' option causes tcpdump to print the `frame
708 control' field, the source and destination addresses, and the packet
709 length. (The `frame control' field governs the interpretation of the
710 rest of the packet. Normal packets (such as those containing IP data‐
711 grams) are `async' packets, with a priority value between 0 and 7; for
712 example, `async4'. Such packets are assumed to contain an 802.2 Logi‐
713 cal Link Control (LLC) packet; the LLC header is printed if it is not
714 an ISO datagram or a so-called SNAP packet.
715 On Token Ring networks, the '-e' option causes tcpdump to print the
717 `access control' and `frame control' fields, the source and destination
718 addresses, and the packet length. As on FDDI networks, packets are as‐
719 sumed to contain an LLC packet. Regardless of whether the '-e' option
720 is specified or not, the source routing information is printed for
721 source-routed packets.
722 On 802.11 networks, the '-e' option causes tcpdump to print the `frame
724 control' fields, all of the addresses in the 802.11 header, and the
725 packet length. As on FDDI networks, packets are assumed to contain an
726 LLC packet.
727 (N.B.: The following description assumes familiarity with the SLIP com‐
729 pression algorithm described in RFC 1144.)
730 On SLIP links, a direction indicator (``I'' for inbound, ``O'' for out‐
732 bound), packet type, and compression information are printed out. The
733 packet type is printed first. The three types are ip, utcp, and ctcp.
734 No further link information is printed for ip packets. For TCP pack‐
735 ets, the connection identifier is printed following the type. If the
736 packet is compressed, its encoded header is printed out. The special
737 cases are printed out as *S+n and *SA+n, where n is the amount by which
738 the sequence number (or sequence number and ack) has changed. If it is
739 not a special case, zero or more changes are printed. A change is in‐
740 dicated by U (urgent pointer), W (window), A (ack), S (sequence num‐
741 ber), and I (packet ID), followed by a delta (+n or -n), or a new value
742 (=n). Finally, the amount of data in the packet and compressed header
743 length are printed.
744 For example, the following line shows an outbound compressed TCP
746 packet, with an implicit connection identifier; the ack has changed by
747 6, the sequence number by 49, and the packet ID by 6; there are 3 bytes
748 of data and 6 bytes of compressed header:
749 O ctcp * A+6 S+49 I+6 3 (6)
750 ARP/RARP Packets
752 ARP/RARP output shows the type of request and its arguments. The for‐
754 mat is intended to be self explanatory. Here is a short sample taken
755 from the start of an `rlogin' from host rtsg to host csam:
756 arp who-has csam tell rtsg
757 arp reply csam is-at CSAM
758 The first line says that rtsg sent an ARP packet asking for the Ether‐
759 net address of internet host csam. Csam replies with its Ethernet ad‐
760 dress (in this example, Ethernet addresses are in caps and internet ad‐
761 dresses in lower case).
762 This would look less redundant if we had done tcpdump -n:
764 arp who-has 128.3.254.6 tell 128.3.254.68
765 arp reply 128.3.254.6 is-at 02:07:01:00:01:c4
766 If we had done tcpdump -e, the fact that the first packet is broadcast
768 and the second is point-to-point would be visible:
769 RTSG Broadcast 0806 64: arp who-has csam tell rtsg
770 CSAM RTSG 0806 64: arp reply csam is-at CSAM
771 For the first packet this says the Ethernet source address is RTSG, the
772 destination is the Ethernet broadcast address, the type field contained
773 hex 0806 (type ETHER_ARP) and the total length was 64 bytes.
774 IPv4 Packets
776 If the link-layer header is not being printed, for IPv4 packets, IP is
778 printed after the time stamp.
779 If the -v flag is specified, information from the IPv4 header is shown
781 in parentheses after the IP or the link-layer header. The general for‐
782 mat of this information is:
783 tos tos, ttl ttl, id id, offset offset, flags [flags], proto proto, length length, options (options)
784 tos is the type of service field; if the ECN bits are non-zero, those
785 are reported as ECT(1), ECT(0), or CE. ttl is the time-to-live; it is
786 not reported if it is zero. id is the IP identification field. offset
787 is the fragment offset field; it is printed whether this is part of a
788 fragmented datagram or not. flags are the MF and DF flags; + is re‐
789 ported if MF is set, and DF is reported if F is set. If neither are
790 set, . is reported. proto is the protocol ID field. length is the to‐
791 tal length field. options are the IP options, if any.
792 Next, for TCP and UDP packets, the source and destination IP addresses
794 and TCP or UDP ports, with a dot between each IP address and its corre‐
795 sponding port, will be printed, with a > separating the source and des‐
796 tination. For other protocols, the addresses will be printed, with a >
797 separating the source and destination. Higher level protocol informa‐
798 tion, if any, will be printed after that.
799 For fragmented IP datagrams, the first fragment contains the higher
801 level protocol header; fragments after the first contain no higher
802 level protocol header. Fragmentation information will be printed only
803 with the -v flag, in the IP header information, as described above.
804 TCP Packets
806 (N.B.:The following description assumes familiarity with the TCP proto‐
808 col described in RFC 793. If you are not familiar with the protocol,
809 this description will not be of much use to you.)
810 The general format of a TCP protocol line is:
812 src > dst: Flags [tcpflags], seq data-seqno, ack ackno, win window, urg urgent, options [opts], length len
813 Src and dst are the source and destination IP addresses and ports.
814 Tcpflags are some combination of S (SYN), F (FIN), P (PSH), R (RST), U
815 (URG), W (CWR), E (ECE) or `.' (ACK), or `none' if no flags are set.
816 Data-seqno describes the portion of sequence space covered by the data
817 in this packet (see example below). Ackno is sequence number of the
818 next data expected the other direction on this connection. Window is
819 the number of bytes of receive buffer space available the other direc‐
820 tion on this connection. Urg indicates there is `urgent' data in the
821 packet. Opts are TCP options (e.g., mss 1024). Len is the length of
822 payload data.
823 Iptype, Src, dst, and flags are always present. The other fields de‐
825 pend on the contents of the packet's TCP protocol header and are output
826 only if appropriate.
827 Here is the opening portion of an rlogin from host rtsg to host csam.
829 IP rtsg.1023 > csam.login: Flags [S], seq 768512:768512, win 4096, opts [mss 1024]
830 IP csam.login > rtsg.1023: Flags [S.], seq, 947648:947648, ack 768513, win 4096, opts [mss 1024]
831 IP rtsg.1023 > csam.login: Flags [.], ack 1, win 4096
832 IP rtsg.1023 > csam.login: Flags [P.], seq 1:2, ack 1, win 4096, length 1
833 IP csam.login > rtsg.1023: Flags [.], ack 2, win 4096
834 IP rtsg.1023 > csam.login: Flags [P.], seq 2:21, ack 1, win 4096, length 19
835 IP csam.login > rtsg.1023: Flags [P.], seq 1:2, ack 21, win 4077, length 1
836 IP csam.login > rtsg.1023: Flags [P.], seq 2:3, ack 21, win 4077, urg 1, length 1
837 IP csam.login > rtsg.1023: Flags [P.], seq 3:4, ack 21, win 4077, urg 1, length 1
838 The first line says that TCP port 1023 on rtsg sent a packet to port
839 login on csam. The S indicates that the SYN flag was set. The packet
840 sequence number was 768512 and it contained no data. (The notation is
841 `first:last' which means `sequence numbers first up to but not includ‐
842 ing last'.) There was no piggy-backed ACK, the available receive win‐
843 dow was 4096 bytes and there was a max-segment-size option requesting
844 an MSS of 1024 bytes.
845 Csam replies with a similar packet except it includes a piggy-backed
847 ACK for rtsg's SYN. Rtsg then ACKs csam's SYN. The `.' means the ACK
848 flag was set. The packet contained no data so there is no data se‐
849 quence number or length. Note that the ACK sequence number is a small
850 integer (1). The first time tcpdump sees a TCP `conversation', it
851 prints the sequence number from the packet. On subsequent packets of
852 the conversation, the difference between the current packet's sequence
853 number and this initial sequence number is printed. This means that
854 sequence numbers after the first can be interpreted as relative byte
855 positions in the conversation's data stream (with the first data byte
856 each direction being `1'). `-S' will override this feature, causing
857 the original sequence numbers to be output.
858 On the 6th line, rtsg sends csam 19 bytes of data (bytes 2 through 20
860 in the rtsg → csam side of the conversation). The PSH flag is set in
861 the packet. On the 7th line, csam says it's received data sent by rtsg
862 up to but not including byte 21. Most of this data is apparently sit‐
863 ting in the socket buffer since csam's receive window has gotten 19
864 bytes smaller. Csam also sends one byte of data to rtsg in this
865 packet. On the 8th and 9th lines, csam sends two bytes of urgent,
866 pushed data to rtsg.
867 If the snapshot was small enough that tcpdump didn't capture the full
869 TCP header, it interprets as much of the header as it can and then re‐
870 ports ``[|tcp]'' to indicate the remainder could not be interpreted.
871 If the header contains a bogus option (one with a length that's either
872 too small or beyond the end of the header), tcpdump reports it as
873 ``[bad opt]'' and does not interpret any further options (since it's
874 impossible to tell where they start). If the header length indicates
875 options are present but the IP datagram length is not long enough for
876 the options to actually be there, tcpdump reports it as ``[bad hdr
877 length]''.
878 Capturing TCP packets with particular flag combinations (SYN-ACK, URG-
880 ACK, etc.)
881 There are 8 bits in the control bits section of the TCP header:
883 CWR | ECE | URG | ACK | PSH | RST | SYN | FIN
885 Let's assume that we want to watch packets used in establishing a TCP
887 connection. Recall that TCP uses a 3-way handshake protocol when it
888 initializes a new connection; the connection sequence with regard to
889 the TCP control bits is
890 1) Caller sends SYN
892 2) Recipient responds with SYN, ACK
893 3) Caller sends ACK
894 Now we're interested in capturing packets that have only the SYN bit
896 set (Step 1). Note that we don't want packets from step 2 (SYN-ACK),
897 just a plain initial SYN. What we need is a correct filter expression
898 for tcpdump.
899 Recall the structure of a TCP header without options:
901 0 15 31
903 -----------------------------------------------------------------
904 | source port | destination port |
905 -----------------------------------------------------------------
906 | sequence number |
907 -----------------------------------------------------------------
908 | acknowledgment number |
909 -----------------------------------------------------------------
910 | HL | rsvd |C|E|U|A|P|R|S|F| window size |
911 -----------------------------------------------------------------
912 | TCP checksum | urgent pointer |
913 -----------------------------------------------------------------
914 A TCP header usually holds 20 octets of data, unless options are
916 present. The first line of the graph contains octets 0 - 3, the second
917 line shows octets 4 - 7 etc.
918 Starting to count with 0, the relevant TCP control bits are contained
920 in octet 13:
921 0 7| 15| 23| 31
923 ----------------|---------------|---------------|----------------
924 | HL | rsvd |C|E|U|A|P|R|S|F| window size |
925 ----------------|---------------|---------------|----------------
926 | | 13th octet | | |
927 Let's have a closer look at octet no. 13:
929 | |
931 |---------------|
932 |C|E|U|A|P|R|S|F|
933 |---------------|
934 |7 5 3 0|
935 These are the TCP control bits we are interested in. We have numbered
937 the bits in this octet from 0 to 7, right to left, so the PSH bit is
938 bit number 3, while the URG bit is number 5.
939 Recall that we want to capture packets with only SYN set. Let's see
941 what happens to octet 13 if a TCP datagram arrives with the SYN bit set
942 in its header:
943 |C|E|U|A|P|R|S|F|
945 |---------------|
946 |0 0 0 0 0 0 1 0|
947 |---------------|
948 |7 6 5 4 3 2 1 0|
949 Looking at the control bits section we see that only bit number 1 (SYN)
951 is set.
952 Assuming that octet number 13 is an 8-bit unsigned integer in network
954 byte order, the binary value of this octet is
955 00000010
957 and its decimal representation is
959 7 6 5 4 3 2 1 0
961 0*2 + 0*2 + 0*2 + 0*2 + 0*2 + 0*2 + 1*2 + 0*2 = 2
962 We're almost done, because now we know that if only SYN is set, the
964 value of the 13th octet in the TCP header, when interpreted as a 8-bit
965 unsigned integer in network byte order, must be exactly 2.
966 This relationship can be expressed as
968 tcp[13] == 2
969 We can use this expression as the filter for tcpdump in order to watch
971 packets which have only SYN set:
972 tcpdump -i xl0 tcp[13] == 2
973 The expression says "let the 13th octet of a TCP datagram have the dec‐
975 imal value 2", which is exactly what we want.
976 Now, let's assume that we need to capture SYN packets, but we don't
978 care if ACK or any other TCP control bit is set at the same time.
979 Let's see what happens to octet 13 when a TCP datagram with SYN-ACK set
980 arrives:
981 |C|E|U|A|P|R|S|F|
983 |---------------|
984 |0 0 0 1 0 0 1 0|
985 |---------------|
986 |7 6 5 4 3 2 1 0|
987 Now bits 1 and 4 are set in the 13th octet. The binary value of octet
989 13 is
990 00010010
992 which translates to decimal
994 7 6 5 4 3 2 1 0
996 0*2 + 0*2 + 0*2 + 1*2 + 0*2 + 0*2 + 1*2 + 0*2 = 18
997 Now we can't just use 'tcp[13] == 18' in the tcpdump filter expression,
999 because that would select only those packets that have SYN-ACK set, but
1000 not those with only SYN set. Remember that we don't care if ACK or any
1001 other control bit is set as long as SYN is set.
1002 In order to achieve our goal, we need to logically AND the binary value
1004 of octet 13 with some other value to preserve the SYN bit. We know
1005 that we want SYN to be set in any case, so we'll logically AND the
1006 value in the 13th octet with the binary value of a SYN:
1007 00010010 SYN-ACK 00000010 SYN
1009 AND 00000010 (we want SYN) AND 00000010 (we want SYN)
1010 -------- --------
1011 = 00000010 = 00000010
1012 We see that this AND operation delivers the same result regardless
1014 whether ACK or another TCP control bit is set. The decimal representa‐
1015 tion of the AND value as well as the result of this operation is 2 (bi‐
1016 nary 00000010), so we know that for packets with SYN set the following
1017 relation must hold true:
1018 ( ( value of octet 13 ) AND ( 2 ) ) == ( 2 )
1020 This points us to the tcpdump filter expression
1022 tcpdump -i xl0 'tcp[13] & 2 == 2'
1023 Some offsets and field values may be expressed as names rather than as
1025 numeric values. For example tcp[13] may be replaced with tcp[tcpflags].
1026 The following TCP flag field values are also available: tcp-fin, tcp-
1027 syn, tcp-rst, tcp-push, tcp-ack, tcp-urg, tcp-ece and tcp-cwr.
1028 This can be demonstrated as:
1030 tcpdump -i xl0 'tcp[tcpflags] & tcp-push != 0'
1031 Note that you should use single quotes or a backslash in the expression
1033 to hide the AND ('&') special character from the shell.
1034 UDP Packets
1036 UDP format is illustrated by this rwho packet:
1038 actinide.who > broadcast.who: udp 84
1039 This says that port who on host actinide sent a UDP datagram to port
1040 who on host broadcast, the Internet broadcast address. The packet con‐
1041 tained 84 bytes of user data.
1042 Some UDP services are recognized (from the source or destination port
1044 number) and the higher level protocol information printed. In particu‐
1045 lar, Domain Name service requests (RFC 1034/1035) and Sun RPC calls
1046 (RFC 1050) to NFS.
1047 TCP or UDP Name Server Requests
1049 (N.B.:The following description assumes familiarity with the Domain
1051 Service protocol described in RFC 1035. If you are not familiar with
1052 the protocol, the following description will appear to be written in
1053 Greek.)
1054 Name server requests are formatted as
1056 src > dst: id op? flags qtype qclass name (len)
1057 h2opolo.1538 > helios.domain: 3+ A? ucbvax.berkeley.edu. (37)
1058 Host h2opolo asked the domain server on helios for an address record
1059 (qtype=A) associated with the name ucbvax.berkeley.edu. The query id
1060 was `3'. The `+' indicates the recursion desired flag was set. The
1061 query length was 37 bytes, excluding the TCP or UDP and IP protocol
1062 headers. The query operation was the normal one, Query, so the op
1063 field was omitted. If the op had been anything else, it would have
1064 been printed between the `3' and the `+'. Similarly, the qclass was
1065 the normal one, C_IN, and omitted. Any other qclass would have been
1066 printed immediately after the `A'.
1067 A few anomalies are checked and may result in extra fields enclosed in
1069 square brackets: If a query contains an answer, authority records or
1070 additional records section, ancount, nscount, or arcount are printed as
1071 `[na]', `[nn]' or `[nau]' where n is the appropriate count. If any of
1072 the response bits are set (AA, RA or rcode) or any of the `must be
1073 zero' bits are set in bytes two and three, `[b2&3=x]' is printed, where
1074 x is the hex value of header bytes two and three.
1075 TCP or UDP Name Server Responses
1077 Name server responses are formatted as
1079 src > dst: id op rcode flags a/n/au type class data (len)
1080 helios.domain > h2opolo.1538: 3 3/3/7 A 128.32.137.3 (273)
1081 helios.domain > h2opolo.1537: 2 NXDomain* 0/1/0 (97)
1082 In the first example, helios responds to query id 3 from h2opolo with 3
1083 answer records, 3 name server records and 7 additional records. The
1084 first answer record is type A (address) and its data is internet ad‐
1085 dress 128.32.137.3. The total size of the response was 273 bytes, ex‐
1086 cluding TCP or UDP and IP headers. The op (Query) and response code
1087 (NoError) were omitted, as was the class (C_IN) of the A record.
1088 In the second example, helios responds to query 2 with a response code
1090 of non-existent domain (NXDomain) with no answers, one name server and
1091 no authority records. The `*' indicates that the authoritative answer
1092 bit was set. Since there were no answers, no type, class or data were
1093 printed.
1094 Other flag characters that might appear are `-' (recursion available,
1096 RA, not set) and `|' (truncated message, TC, set). If the `question'
1097 section doesn't contain exactly one entry, `[nq]' is printed.
1098 SMB/CIFS decoding
1100 tcpdump now includes fairly extensive SMB/CIFS/NBT decoding for data on
1102 UDP/137, UDP/138 and TCP/139. Some primitive decoding of IPX and Net‐
1103 BEUI SMB data is also done.
1104 By default a fairly minimal decode is done, with a much more detailed
1106 decode done if -v is used. Be warned that with -v a single SMB packet
1107 may take up a page or more, so only use -v if you really want all the
1108 gory details.
1109 For information on SMB packet formats and what all the fields mean see
1111 https://download.samba.org/pub/samba/specs/ and other online resources.
1112 The SMB patches were written by Andrew Tridgell (tridge@samba.org).
1113 NFS Requests and Replies
1115 Sun NFS (Network File System) requests and replies are printed as:
1117 src.sport > dst.nfs: NFS request xid xid len op args
1118 src.nfs > dst.dport: NFS reply xid xid reply stat len op results
1119 sushi.1023 > wrl.nfs: NFS request xid 26377
1120 112 readlink fh 21,24/10.73165
1121 wrl.nfs > sushi.1023: NFS reply xid 26377
1122 reply ok 40 readlink "../var"
1123 sushi.1022 > wrl.nfs: NFS request xid 8219
1124 144 lookup fh 9,74/4096.6878 "xcolors"
1125 wrl.nfs > sushi.1022: NFS reply xid 8219
1126 reply ok 128 lookup fh 9,74/4134.3150
1127 In the first line, host sushi sends a transaction with id 26377 to wrl.
1128 The request was 112 bytes, excluding the UDP and IP headers. The oper‐
1129 ation was a readlink (read symbolic link) on file handle (fh)
1130 21,24/10.731657119. (If one is lucky, as in this case, the file handle
1131 can be interpreted as a major,minor device number pair, followed by the
1132 inode number and generation number.) In the second line, wrl replies
1133 `ok' with the same transaction id and the contents of the link.
1134 In the third line, sushi asks (using a new transaction id) wrl to
1136 lookup the name `xcolors' in directory file 9,74/4096.6878. In the
1137 fourth line, wrl sends a reply with the respective transaction id.
1138 Note that the data printed depends on the operation type. The format
1140 is intended to be self explanatory if read in conjunction with an NFS
1141 protocol spec. Also note that older versions of tcpdump printed NFS
1142 packets in a slightly different format: the transaction id (xid) would
1143 be printed instead of the non-NFS port number of the packet.
1144 If the -v (verbose) flag is given, additional information is printed.
1146 For example:
1147 sushi.1023 > wrl.nfs: NFS request xid 79658
1148 148 read fh 21,11/12.195 8192 bytes @ 24576
1149 wrl.nfs > sushi.1023: NFS reply xid 79658
1150 reply ok 1472 read REG 100664 ids 417/0 sz 29388
1151 (-v also prints the IP header TTL, ID, length, and fragmentation
1152 fields, which have been omitted from this example.) In the first line,
1153 sushi asks wrl to read 8192 bytes from file 21,11/12.195, at byte off‐
1154 set 24576. Wrl replies `ok'; the packet shown on the second line is
1155 the first fragment of the reply, and hence is only 1472 bytes long (the
1156 other bytes will follow in subsequent fragments, but these fragments do
1157 not have NFS or even UDP headers and so might not be printed, depending
1158 on the filter expression used). Because the -v flag is given, some of
1159 the file attributes (which are returned in addition to the file data)
1160 are printed: the file type (``REG'', for regular file), the file mode
1161 (in octal), the UID and GID, and the file size.
1162 If the -v flag is given more than once, even more details are printed.
1164 NFS reply packets do not explicitly identify the RPC operation. In‐
1166 stead, tcpdump keeps track of ``recent'' requests, and matches them to
1167 the replies using the transaction ID. If a reply does not closely fol‐
1168 low the corresponding request, it might not be parsable.
1169 AFS Requests and Replies
1171 Transarc AFS (Andrew File System) requests and replies are printed as:
1173 src.sport > dst.dport: rx packet-type
1175 src.sport > dst.dport: rx packet-type service call call-name args
1176 src.sport > dst.dport: rx packet-type service reply call-name args
1177 elvis.7001 > pike.afsfs:
1178 rx data fs call rename old fid 536876964/1/1 ".newsrc.new"
1179 new fid 536876964/1/1 ".newsrc"
1180 pike.afsfs > elvis.7001: rx data fs reply rename
1181 In the first line, host elvis sends a RX packet to pike. This was a RX
1182 data packet to the fs (fileserver) service, and is the start of an RPC
1183 call. The RPC call was a rename, with the old directory file id of
1184 536876964/1/1 and an old filename of `.newsrc.new', and a new directory
1185 file id of 536876964/1/1 and a new filename of `.newsrc'. The host
1186 pike responds with a RPC reply to the rename call (which was success‐
1187 ful, because it was a data packet and not an abort packet).
1188 In general, all AFS RPCs are decoded at least by RPC call name. Most
1190 AFS RPCs have at least some of the arguments decoded (generally only
1191 the `interesting' arguments, for some definition of interesting).
1192 The format is intended to be self-describing, but it will probably not
1194 be useful to people who are not familiar with the workings of AFS and
1195 RX.
1196 If the -v (verbose) flag is given twice, acknowledgement packets and
1198 additional header information is printed, such as the RX call ID, call
1199 number, sequence number, serial number, and the RX packet flags.
1200 If the -v flag is given twice, additional information is printed, such
1202 as the RX call ID, serial number, and the RX packet flags. The MTU ne‐
1203 gotiation information is also printed from RX ack packets.
1204 If the -v flag is given three times, the security index and service id
1206 are printed.
1207 Error codes are printed for abort packets, with the exception of Ubik
1209 beacon packets (because abort packets are used to signify a yes vote
1210 for the Ubik protocol).
1211 AFS reply packets do not explicitly identify the RPC operation. In‐
1213 stead, tcpdump keeps track of ``recent'' requests, and matches them to
1214 the replies using the call number and service ID. If a reply does not
1215 closely follow the corresponding request, it might not be parsable.
1216 KIP AppleTalk (DDP in UDP)
1219 AppleTalk DDP packets encapsulated in UDP datagrams are de-encapsulated
1221 and dumped as DDP packets (i.e., all the UDP header information is dis‐
1222 carded). The file /etc/atalk.names is used to translate AppleTalk net
1223 and node numbers to names. Lines in this file have the form
1224 number name
1225 1.254 ether
1227 16.1 icsd-net
1228 1.254.110 ace
1229 The first two lines give the names of AppleTalk networks. The third
1230 line gives the name of a particular host (a host is distinguished from
1231 a net by the 3rd octet in the number - a net number must have two
1232 octets and a host number must have three octets.) The number and name
1233 should be separated by whitespace (blanks or tabs). The
1234 /etc/atalk.names file may contain blank lines or comment lines (lines
1235 starting with a `#').
1236 AppleTalk addresses are printed in the form
1238 net.host.port
1239 144.1.209.2 > icsd-net.112.220
1241 office.2 > icsd-net.112.220
1242 jssmag.149.235 > icsd-net.2
1243 (If the /etc/atalk.names doesn't exist or doesn't contain an entry for
1244 some AppleTalk host/net number, addresses are printed in numeric form.)
1245 In the first example, NBP (DDP port 2) on net 144.1 node 209 is sending
1246 to whatever is listening on port 220 of net icsd node 112. The second
1247 line is the same except the full name of the source node is known (`of‐
1248 fice'). The third line is a send from port 235 on net jssmag node 149
1249 to broadcast on the icsd-net NBP port (note that the broadcast address
1250 (255) is indicated by a net name with no host number - for this reason
1251 it's a good idea to keep node names and net names distinct in
1252 /etc/atalk.names).
1253 NBP (name binding protocol) and ATP (AppleTalk transaction protocol)
1255 packets have their contents interpreted. Other protocols just dump the
1256 protocol name (or number if no name is registered for the protocol) and
1257 packet size.
1258 NBP packets are formatted like the following examples:
1260 icsd-net.112.220 > jssmag.2: nbp-lkup 190: "=:LaserWriter@*"
1261 jssmag.209.2 > icsd-net.112.220: nbp-reply 190: "RM1140:LaserWriter@*" 250
1262 techpit.2 > icsd-net.112.220: nbp-reply 190: "techpit:LaserWriter@*" 186
1263 The first line is a name lookup request for laserwriters sent by net
1264 icsd host 112 and broadcast on net jssmag. The nbp id for the lookup
1265 is 190. The second line shows a reply for this request (note that it
1266 has the same id) from host jssmag.209 saying that it has a laserwriter
1267 resource named "RM1140" registered on port 250. The third line is an‐
1268 other reply to the same request saying host techpit has laserwriter
1269 "techpit" registered on port 186.
1270 ATP packet formatting is demonstrated by the following example:
1272 jssmag.209.165 > helios.132: atp-req 12266<0-7> 0xae030001
1273 helios.132 > jssmag.209.165: atp-resp 12266:0 (512) 0xae040000
1274 helios.132 > jssmag.209.165: atp-resp 12266:1 (512) 0xae040000
1275 helios.132 > jssmag.209.165: atp-resp 12266:2 (512) 0xae040000
1276 helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
1277 helios.132 > jssmag.209.165: atp-resp 12266:4 (512) 0xae040000
1278 helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
1279 helios.132 > jssmag.209.165: atp-resp 12266:6 (512) 0xae040000
1280 helios.132 > jssmag.209.165: atp-resp*12266:7 (512) 0xae040000
1281 jssmag.209.165 > helios.132: atp-req 12266<3,5> 0xae030001
1282 helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
1283 helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
1284 jssmag.209.165 > helios.132: atp-rel 12266<0-7> 0xae030001
1285 jssmag.209.133 > helios.132: atp-req* 12267<0-7> 0xae030002
1286 Jssmag.209 initiates transaction id 12266 with host helios by request‐
1287 ing up to 8 packets (the `<0-7>'). The hex number at the end of the
1288 line is the value of the `userdata' field in the request.
1289 Helios responds with 8 512-byte packets. The `:digit' following the
1291 transaction id gives the packet sequence number in the transaction and
1292 the number in parens is the amount of data in the packet, excluding the
1293 ATP header. The `*' on packet 7 indicates that the EOM bit was set.
1294 Jssmag.209 then requests that packets 3 & 5 be retransmitted. Helios
1296 resends them then jssmag.209 releases the transaction. Finally, jss‐
1297 mag.209 initiates the next request. The `*' on the request indicates
1298 that XO (`exactly once') was not set.
1299
1302 The TCP flag names tcp-ece and tcp-cwr became available when linking
1303 with libpcap 1.9.0 or later.
1304
1307 stty(1), pcap(3PCAP), bpf(4), nit(4P), pcap-savefile(5),
1308 pcap-filter(7), pcap-tstamp(7)
1309 https://www.iana.org/assignments/media-types/applica‐
1311 tion/vnd.tcpdump.pcap
1312
1314 The original authors are:
1315 Van Jacobson, Craig Leres and Steven McCanne, all of the Lawrence
1317 Berkeley National Laboratory, University of California, Berkeley, CA.
1318 It is currently maintained by The Tcpdump Group.
1320 The current version is available via HTTPS:
1322 https://www.tcpdump.org/
1324 The original distribution is available via anonymous ftp:
1326 ftp://ftp.ee.lbl.gov/old/tcpdump.tar.Z
1328 IPv6/IPsec support is added by WIDE/KAME project. This program uses
1330 OpenSSL/LibreSSL, under specific configurations.
1331
1333 To report a security issue please send an e-mail to
1334 security@tcpdump.org.
1335 To report bugs and other problems, contribute patches, request a fea‐
1337 ture, provide generic feedback etc. please see the file CONTRIBUTING.md
1338 in the tcpdump source tree root.
1339 NIT doesn't let you watch your own outbound traffic, BPF will. We rec‐
1341 ommend that you use the latter.
1342 On Linux systems with 2.0[.x] kernels:
1344 packets on the loopback device will be seen twice;
1346 packet filtering cannot be done in the kernel, so that all pack‐
1348 ets must be copied from the kernel in order to be filtered in
1349 user mode;
1350 all of a packet, not just the part that's within the snapshot
1352 length, will be copied from the kernel (the 2.0[.x] packet cap‐
1353 ture mechanism, if asked to copy only part of a packet to
1354 userspace, will not report the true length of the packet; this
1355 would cause most IP packets to get an error from tcpdump);
1356 capturing on some PPP devices won't work correctly.
1358 We recommend that you upgrade to a 2.2 or later kernel.
1360 Some attempt should be made to reassemble IP fragments or, at least to
1362 compute the right length for the higher level protocol.
1363 Name server inverse queries are not dumped correctly: the (empty) ques‐
1365 tion section is printed rather than real query in the answer section.
1366 Some believe that inverse queries are themselves a bug and prefer to
1367 fix the program generating them rather than tcpdump.
1368 A packet trace that crosses a daylight savings time change will give
1370 skewed time stamps (the time change is ignored).
1371 Filter expressions on fields other than those in Token Ring headers
1373 will not correctly handle source-routed Token Ring packets.
1374 Filter expressions on fields other than those in 802.11 headers will
1376 not correctly handle 802.11 data packets with both To DS and From DS
1377 set.
1378 ip6 proto should chase header chain, but at this moment it does not.
1380 ip6 protochain is supplied for this behavior.
1381 Arithmetic expression against transport layer headers, like tcp[0],
1383 does not work against IPv6 packets. It only looks at IPv4 packets.
1384 12 March 2023 TCPDUMP(8)