1 .\" @(#) $Header: tcpdump.1,v 1.67 97/06/30 16:31:50 leres Exp $ (LBL)
3 .\" Copyright (c) 1987, 1988, 1989, 1990, 1991, 1992, 1994, 1995, 1996, 1997
4 .\" The Regents of the University of California. All rights reserved.
5 .\" All rights reserved.
7 .\" Redistribution and use in source and binary forms, with or without
8 .\" modification, are permitted provided that: (1) source code distributions
9 .\" retain the above copyright notice and this paragraph in its entirety, (2)
10 .\" distributions including binary code include the above copyright notice and
11 .\" this paragraph in its entirety in the documentation or other materials
12 .\" provided with the distribution, and (3) all advertising materials mentioning
13 .\" features or use of this software display the following acknowledgement:
14 .\" ``This product includes software developed by the University of California,
15 .\" Lawrence Berkeley Laboratory and its contributors.'' Neither the name of
16 .\" the University nor the names of its contributors may be used to endorse
17 .\" or promote products derived from this software without specific prior
18 .\" written permission.
19 .\" THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED
20 .\" WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
21 .\" MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
23 .TH TCPDUMP 1 "30 June 1997"
25 tcpdump \- dump traffic on a network
68 \fITcpdump\fP prints out the headers of packets on a network interface
69 that match the boolean \fIexpression\fP.
71 .B Under SunOS with nit or bpf:
74 you must have read access to
78 .B Under Solaris with dlpi:
79 You must have read access to the network pseudo device, e.g.
81 .B Under HP-UX with dlpi:
82 You must be root or it must be installed setuid to root.
83 .B Under IRIX with snoop:
84 You must be root or it must be installed setuid to root.
86 You must be root or it must be installed setuid to root.
87 .B Under Ultrix and Digital UNIX:
88 Once the super-user has enabled promiscuous-mode operation using
93 You must have read access to
98 Attempt to convert network and broadcast addresses to names.
101 Exit after receiving \fIcount\fP packets.
104 Dump the compiled packet-matching code in a human readable form to
105 standard output and stop.
108 Dump packet-matching code as a
113 Dump packet-matching code as decimal numbers (preceded with a count).
116 Print the link-level header on each dump line.
119 Print `foreign' internet addresses numerically rather than symbolically
120 (this option is intended to get around serious brain damage in
121 Sun's yp server \(em usually it hangs forever translating non-local
125 Use \fIfile\fP as input for the filter expression.
126 An additional expression given on the command line is ignored.
129 Listen on \fIinterface\fP.
130 If unspecified, \fItcpdump\fP searches the system interface list for the
131 lowest numbered, configured up interface (excluding loopback).
132 Ties are broken by choosing the earliest match.
135 Make stdout line buffered. Useful if you want to see the data
136 while capturing it. E.g.,
138 ``tcpdump\ \ \-l\ \ |\ \ tee dat'' or
139 ``tcpdump\ \ \-l \ \ > dat\ \ &\ \ tail\ \ \-f\ \ dat''.
142 Don't convert addresses (i.e., host addresses, port numbers, etc.) to names.
145 Don't print domain name qualification of host names. E.g.,
146 if you give this flag then \fItcpdump\fP will print ``nic''
147 instead of ``nic.ddn.mil''.
150 Do not run the packet-matching code optimizer. This is useful only
151 if you suspect a bug in the optimizer.
154 \fIDon't\fP put the interface
155 into promiscuous mode. Note that the interface might be in promiscuous
156 mode for some other reason; hence, `-p' cannot be used as an abbreviation for
157 `ether host {local-hw-addr} or ether broadcast'.
160 Quick (quiet?) output. Print less protocol information so output
164 Read packets from \fIfile\fR (which was created with the -w option).
165 Standard input is used if \fIfile\fR is ``-''.
168 Snarf \fIsnaplen\fP bytes of data from each packet rather than the
169 default of 68 (with SunOS's NIT, the minimum is actually 96).
170 68 bytes is adequate for IP, ICMP, TCP
171 and UDP but may truncate protocol information from name server and NFS
172 packets (see below). Packets truncated because of a limited snapshot
173 are indicated in the output with ``[|\fIproto\fP]'', where \fIproto\fP
174 is the name of the protocol level at which the truncation has occurred.
175 Note that taking larger snapshots both increases
176 the amount of time it takes to process packets and, effectively,
177 decreases the amount of packet buffering. This may cause packets to be
178 lost. You should limit \fIsnaplen\fP to the smallest number that will
179 capture the protocol information you're interested in.
182 Force packets selected by "\fIexpression\fP" to be interpreted the
183 specified \fItype\fR. Currently known types are
184 \fBrpc\fR (Remote Procedure Call),
185 \fBrtp\fR (Real-Time Applications protocol),
186 \fBrtcp\fR (Real-Time Applications control protocol),
187 \fBvat\fR (Visual Audio Tool),
189 \fBwb\fR (distributed White Board).
192 Print absolute, rather than relative, TCP sequence numbers.
195 \fIDon't\fP print a timestamp on each dump line.
198 Print an unformatted timestamp on each dump line.
201 (Slightly more) verbose output. For example, the time to live
202 and type of service information in an IP packet is printed.
205 Even more verbose output. For example, additional fields are
206 printed from NFS reply packets.
209 Write the raw packets to \fIfile\fR rather than parsing and printing
210 them out. They can later be printed with the \-r option.
211 Standard output is used if \fIfile\fR is ``-''.
214 Print each packet (minus its link level header) in hex.
215 The smaller of the entire packet or
217 bytes will be printed.
218 .IP "\fI expression\fP"
220 selects which packets will be dumped. If no \fIexpression\fP
221 is given, all packets on the net will be dumped. Otherwise,
222 only packets for which \fIexpression\fP is `true' will be dumped.
224 The \fIexpression\fP consists of one or more
226 Primitives usually consist of an
228 (name or number) preceded by one or more qualifiers. There are three
229 different kinds of qualifier:
231 qualifiers say what kind of thing the id name or number refers to.
237 E.g., `host foo', `net 128.3', `port 20'. If there is no type
242 qualifiers specify a particular transfer direction to and/or from
244 Possible directions are
251 E.g., `src foo', `dst net 128.3', `src or dst port ftp-data'. If
252 there is no dir qualifier,
255 For `null' link layers (i.e. point to point protocols such as slip) the
259 qualifiers can be used to specify a desired direction.
261 qualifiers restrict the match to a particular protocol. Possible
279 E.g., `ether src foo', `arp net 128.3', `tcp port 21'. If there is
280 no proto qualifier, all protocols consistent with the type are
281 assumed. E.g., `src foo' means `(ip or arp or rarp) src foo'
282 (except the latter is not legal syntax), `net bar' means `(ip or
283 arp or rarp) net bar' and `port 53' means `(tcp or udp) port 53'.
285 [`fddi' is actually an alias for `ether'; the parser treats them
286 identically as meaning ``the data link level used on the specified
287 network interface.'' FDDI headers contain Ethernet-like source
288 and destination addresses, and often contain Ethernet-like packet
289 types, so you can filter on these FDDI fields just as with the
290 analogous Ethernet fields. FDDI headers also contain other fields,
291 but you cannot name them explicitly in a filter expression.]
293 In addition to the above, there are some special `primitive' keywords
294 that don't follow the pattern:
299 and arithmetic expressions. All of these are described below.
301 More complex filter expressions are built up by using the words
306 to combine primitives. E.g., `host foo and not port ftp and not port ftp-data'.
307 To save typing, identical qualifier lists can be omitted. E.g.,
308 `tcp dst port ftp or ftp-data or domain' is exactly the same as
309 `tcp dst port ftp or tcp dst port ftp-data or tcp dst port domain'.
311 Allowable primitives are:
312 .IP "\fBdst host \fIhost\fR"
313 True if the IP destination field of the packet is \fIhost\fP,
314 which may be either an address or a name.
315 .IP "\fBsrc host \fIhost\fR"
316 True if the IP source field of the packet is \fIhost\fP.
317 .IP "\fBhost \fIhost\fP
318 True if either the IP source or destination of the packet is \fIhost\fP.
319 Any of the above host expressions can be prepended with the keywords,
320 \fBip\fP, \fBarp\fP, or \fBrarp\fP as in:
323 \fBip host \fIhost\fR
326 which is equivalent to:
329 \fBether proto \fI\\ip\fB and host \fIhost\fR
332 If \fIhost\fR is a name with multiple IP addresses, each address will
333 be checked for a match.
334 .IP "\fBether dst \fIehost\fP
335 True if the ethernet destination address is \fIehost\fP. \fIEhost\fP
336 may be either a name from /etc/ethers or a number (see
339 .IP "\fBether src \fIehost\fP
340 True if the ethernet source address is \fIehost\fP.
341 .IP "\fBether host \fIehost\fP
342 True if either the ethernet source or destination address is \fIehost\fP.
343 .IP "\fBgateway\fP \fIhost\fP
344 True if the packet used \fIhost\fP as a gateway. I.e., the ethernet
345 source or destination address was \fIhost\fP but neither the IP source
346 nor the IP destination was \fIhost\fP. \fIHost\fP must be a name and
347 must be found in both /etc/hosts and /etc/ethers. (An equivalent
351 \fBether host \fIehost \fBand not host \fIhost\fR
354 which can be used with either names or numbers for \fIhost / ehost\fP.)
355 .IP "\fBdst net \fInet\fR"
356 True if the IP destination address of the packet has a network
357 number of \fInet\fP. \fINet\fP may be either a name from /etc/networks
358 or a network number (see \fInetworks(4)\fP for details).
359 .IP "\fBsrc net \fInet\fR"
360 True if the IP source address of the packet has a network
362 .IP "\fBnet \fInet\fR"
363 True if either the IP source or destination address of the packet has a network
365 .IP "\fBnet \fInet\fR \fBmask \fImask\fR"
366 True if the IP address matches \fInet\fR with the specific netmask.
367 May be qualified with \fBsrc\fR or \fBdst\fR.
368 .IP "\fBnet \fInet\fR/\fIlen\fR"
369 True if the IP address matches \fInet\fR a netmask \fIlen\fR bits wide.
370 May be qualified with \fBsrc\fR or \fBdst\fR.
371 .IP "\fBdst port \fIport\fR"
372 True if the packet is ip/tcp or ip/udp and has a
373 destination port value of \fIport\fP.
374 The \fIport\fP can be a number or a name used in /etc/services (see
378 If a name is used, both the port
379 number and protocol are checked. If a number or ambiguous name is used,
380 only the port number is checked (e.g., \fBdst port 513\fR will print both
381 tcp/login traffic and udp/who traffic, and \fBport domain\fR will print
382 both tcp/domain and udp/domain traffic).
383 .IP "\fBsrc port \fIport\fR"
384 True if the packet has a source port value of \fIport\fP.
385 .IP "\fBport \fIport\fR"
386 True if either the source or destination port of the packet is \fIport\fP.
387 Any of the above port expressions can be prepended with the keywords,
388 \fBtcp\fP or \fBudp\fP, as in:
391 \fBtcp src port \fIport\fR
394 which matches only tcp packets whose source port is \fIport\fP.
395 .IP "\fBless \fIlength\fR"
396 True if the packet has a length less than or equal to \fIlength\fP.
397 This is equivalent to:
400 \fBlen <= \fIlength\fP.
403 .IP "\fBgreater \fIlength\fR"
404 True if the packet has a length greater than or equal to \fIlength\fP.
405 This is equivalent to:
408 \fBlen >= \fIlength\fP.
411 .IP "\fBip proto \fIprotocol\fR"
412 True if the packet is an ip packet (see
414 of protocol type \fIprotocol\fP.
415 \fIProtocol\fP can be a number or one of the names
416 \fIicmp\fP, \fIigrp\fP, \fIudp\fP, \fInd\fP, or \fItcp\fP.
417 Note that the identifiers \fItcp\fP, \fIudp\fP, and \fIicmp\fP are also
418 keywords and must be escaped via backslash (\\), which is \\\\ in the C-shell.
419 .IP "\fBether broadcast\fR"
420 True if the packet is an ethernet broadcast packet. The \fIether\fP
422 .IP "\fBip broadcast\fR"
423 True if the packet is an IP broadcast packet. It checks for both
424 the all-zeroes and all-ones broadcast conventions, and looks up
425 the local subnet mask.
426 .IP "\fBether multicast\fR"
427 True if the packet is an ethernet multicast packet. The \fIether\fP
429 This is shorthand for `\fBether[0] & 1 != 0\fP'.
430 .IP "\fBip multicast\fR"
431 True if the packet is an IP multicast packet.
432 .IP "\fBether proto \fIprotocol\fR"
433 True if the packet is of ether type \fIprotocol\fR.
434 \fIProtocol\fP can be a number or a name like
435 \fIip\fP, \fIarp\fP, or \fIrarp\fP.
436 Note these identifiers are also keywords
437 and must be escaped via backslash (\\).
438 [In the case of FDDI (e.g., `\fBfddi protocol arp\fR'), the
439 protocol identification comes from the 802.2 Logical Link Control
440 (LLC) header, which is usually layered on top of the FDDI header.
441 \fITcpdump\fP assumes, when filtering on the protocol identifier,
442 that all FDDI packets include an LLC header, and that the LLC header
443 is in so-called SNAP format.]
444 .IP "\fBdecnet src \fIhost\fR"
445 True if the DECNET source address is
447 which may be an address of the form ``10.123'', or a DECNET host
448 name. [DECNET host name support is only available on Ultrix systems
449 that are configured to run DECNET.]
450 .IP "\fBdecnet dst \fIhost\fR"
451 True if the DECNET destination address is
453 .IP "\fBdecnet host \fIhost\fR"
454 True if either the DECNET source or destination address is
456 .IP "\fBip\fR, \fBarp\fR, \fBrarp\fR, \fBdecnet\fR, \fBiso\fR"
460 \fBether proto \fIp\fR
463 where \fIp\fR is one of the above protocols.
464 .IP "\fBlat\fR, \fBmoprc\fR, \fBmopdl\fR"
468 \fBether proto \fIp\fR
471 where \fIp\fR is one of the above protocols.
473 \fItcpdump\fP does not currently know how to parse these protocols.
474 .IP "\fBtcp\fR, \fBudp\fR, \fBicmp\fR"
481 where \fIp\fR is one of the above protocols.
482 .IP "\fBesis\fR, \fBisis\fR"
489 where \fIp\fR is one of the above protocols.
490 Note that \fItcpdump\fR does an incomplete job of parsing these protocols.
491 .IP "\fIexpr relop expr\fR"
492 True if the relation holds, where \fIrelop\fR is one of >, <, >=, <=, =, !=,
493 and \fIexpr\fR is an arithmetic expression composed of integer constants
494 (expressed in standard C syntax), the normal binary operators
495 [+, -, *, /, &, |], a length operator, and special packet data accessors.
497 data inside the packet, use the following syntax:
500 \fIproto\fB [ \fIexpr\fB : \fIsize\fB ]\fR
503 \fIProto\fR is one of \fBether, fddi,
504 ip, arp, rarp, tcp, udp, \fRor \fBicmp\fR, and
505 indicates the protocol layer for the index operation.
506 The byte offset, relative to the indicated protocol layer, is
508 \fISize\fR is optional and indicates the number of bytes in the
509 field of interest; it can be either one, two, or four, and defaults to one.
510 The length operator, indicated by the keyword \fBlen\fP, gives the
511 length of the packet.
513 For example, `\fBether[0] & 1 != 0\fP' catches all multicast traffic.
514 The expression `\fBip[0] & 0xf != 5\fP'
515 catches all IP packets with options. The expression
516 `\fBip[6:2] & 0x1fff = 0\fP'
517 catches only unfragmented datagrams and frag zero of fragmented datagrams.
518 This check is implicitly applied to the \fBtcp\fP and \fBudp\fP
520 For instance, \fBtcp[0]\fP always means the first
521 byte of the TCP \fIheader\fP, and never means the first byte of an
522 intervening fragment.
524 Primitives may be combined using:
526 A parenthesized group of primitives and operators
527 (parentheses are special to the Shell and must be escaped).
529 Negation (`\fB!\fP' or `\fBnot\fP').
531 Concatenation (`\fB&&\fP' or `\fBand\fP').
533 Alternation (`\fB||\fP' or `\fBor\fP').
535 Negation has highest precedence.
536 Alternation and concatenation have equal precedence and associate
537 left to right. Note that explicit \fBand\fR tokens, not juxtaposition,
538 are now required for concatenation.
540 If an identifier is given without a keyword, the most recent keyword
545 \fBnot host vs and ace\fR
551 \fBnot host vs and host ace\fR
554 which should not be confused with
557 \fBnot ( host vs or ace )\fR
561 Expression arguments can be passed to tcpdump as either a single argument
562 or as multiple arguments, whichever is more convenient.
563 Generally, if the expression contains Shell metacharacters, it is
564 easier to pass it as a single, quoted argument.
565 Multiple arguments are concatenated with spaces before being parsed.
568 To print all packets arriving at or departing from \fIsundown\fP:
571 \fBtcpdump host sundown\fP
575 To print traffic between \fIhelios\fR and either \fIhot\fR or \fIace\fR:
578 \fBtcpdump host helios and \\( hot or ace \\)\fP
582 To print all IP packets between \fIace\fR and any host except \fIhelios\fR:
585 \fBtcpdump ip host ace and not helios\fP
589 To print all traffic between local hosts and hosts at Berkeley:
593 tcpdump net ucb-ether
597 To print all ftp traffic through internet gateway \fIsnup\fP:
598 (note that the expression is quoted to prevent the shell from
599 (mis-)interpreting the parentheses):
603 tcpdump 'gateway snup and (port ftp or ftp-data)'
607 To print traffic neither sourced from nor destined for local hosts
608 (if you gateway to one other net, this stuff should never make it
609 onto your local net).
613 tcpdump ip and not net \fIlocalnet\fP
617 To print the start and end packets (the SYN and FIN packets) of each
618 TCP conversation that involves a non-local host.
622 tcpdump 'tcp[13] & 3 != 0 and not src and dst net \fIlocalnet\fP'
626 To print IP packets longer than 576 bytes sent through gateway \fIsnup\fP:
630 tcpdump 'gateway snup and ip[2:2] > 576'
634 To print IP broadcast or multicast packets that were
636 sent via ethernet broadcast or multicast:
640 tcpdump 'ether[0] & 1 = 0 and ip[16] >= 224'
644 To print all ICMP packets that are not echo requests/replies (i.e., not
649 tcpdump 'icmp[0] != 8 and icmp[0] != 0'
654 The output of \fItcpdump\fP is protocol dependent. The following
655 gives a brief description and examples of most of the formats.
663 If the '-e' option is given, the link level header is printed out.
664 On ethernets, the source and destination addresses, protocol,
665 and packet length are printed.
667 On FDDI networks, the '-e' option causes \fItcpdump\fP to print
668 the `frame control' field, the source and destination addresses,
669 and the packet length. (The `frame control' field governs the
670 interpretation of the rest of the packet. Normal packets (such
671 as those containing IP datagrams) are `async' packets, with a priority
672 value between 0 and 7; for example, `\fBasync4\fR'. Such packets
673 are assumed to contain an 802.2 Logical Link Control (LLC) packet;
674 the LLC header is printed if it is \fInot\fR an ISO datagram or a
675 so-called SNAP packet.
677 \fI(N.B.: The following description assumes familiarity with
678 the SLIP compression algorithm described in RFC-1144.)\fP
680 On SLIP links, a direction indicator (``I'' for inbound, ``O'' for outbound),
681 packet type, and compression information are printed out.
682 The packet type is printed first.
683 The three types are \fIip\fP, \fIutcp\fP, and \fIctcp\fP.
684 No further link information is printed for \fIip\fR packets.
685 For TCP packets, the connection identifier is printed following the type.
686 If the packet is compressed, its encoded header is printed out.
687 The special cases are printed out as
688 \fB*S+\fIn\fR and \fB*SA+\fIn\fR, where \fIn\fR is the amount by which
689 the sequence number (or sequence number and ack) has changed.
690 If it is not a special case,
691 zero or more changes are printed.
692 A change is indicated by U (urgent pointer), W (window), A (ack),
693 S (sequence number), and I (packet ID), followed by a delta (+n or -n),
695 Finally, the amount of data in the packet and compressed header length
698 For example, the following line shows an outbound compressed TCP packet,
699 with an implicit connection identifier; the ack has changed by 6,
700 the sequence number by 49, and the packet ID by 6; there are 3 bytes of
701 data and 6 bytes of compressed header:
704 \fBO ctcp * A+6 S+49 I+6 3 (6)\fP
710 Arp/rarp output shows the type of request and its arguments. The
711 format is intended to be self explanatory.
712 Here is a short sample taken from the start of an `rlogin' from
713 host \fIrtsg\fP to host \fIcsam\fP:
717 \f(CWarp who-has csam tell rtsg
718 arp reply csam is-at CSAM\fP
722 The first line says that rtsg sent an arp packet asking
723 for the ethernet address of internet host csam. Csam
724 replies with its ethernet address (in this example, ethernet addresses
725 are in caps and internet addresses in lower case).
727 This would look less redundant if we had done \fBtcpdump \-n\fP:
731 \f(CWarp who-has 128.3.254.6 tell 128.3.254.68
732 arp reply 128.3.254.6 is-at 02:07:01:00:01:c4\fP
736 If we had done \fBtcpdump \-e\fP, the fact that the first packet is
737 broadcast and the second is point-to-point would be visible:
741 \f(CWRTSG Broadcast 0806 64: arp who-has csam tell rtsg
742 CSAM RTSG 0806 64: arp reply csam is-at CSAM\fP
746 For the first packet this says the ethernet source address is RTSG, the
747 destination is the ethernet broadcast address, the type field
748 contained hex 0806 (type ETHER_ARP) and the total length was 64 bytes.
752 \fI(N.B.:The following description assumes familiarity with
753 the TCP protocol described in RFC-793. If you are not familiar
754 with the protocol, neither this description nor tcpdump will
755 be of much use to you.)\fP
757 The general format of a tcp protocol line is:
761 \fIsrc > dst: flags data-seqno ack window urgent options\fP
765 \fISrc\fP and \fIdst\fP are the source and destination IP
766 addresses and ports. \fIFlags\fP are some combination of S (SYN),
767 F (FIN), P (PUSH) or R (RST) or a single `.' (no flags).
768 \fIData-seqno\fP describes the portion of sequence space covered
769 by the data in this packet (see example below).
770 \fIAck\fP is sequence number of the next data expected the other
771 direction on this connection.
772 \fIWindow\fP is the number of bytes of receive buffer space available
773 the other direction on this connection.
774 \fIUrg\fP indicates there is `urgent' data in the packet.
775 \fIOptions\fP are tcp options enclosed in angle brackets (e.g., <mss 1024>).
777 \fISrc, dst\fP and \fIflags\fP are always present. The other fields
778 depend on the contents of the packet's tcp protocol header and
779 are output only if appropriate.
781 Here is the opening portion of an rlogin from host \fIrtsg\fP to
786 \s-2\f(CWrtsg.1023 > csam.login: S 768512:768512(0) win 4096 <mss 1024>
787 csam.login > rtsg.1023: S 947648:947648(0) ack 768513 win 4096 <mss 1024>
788 rtsg.1023 > csam.login: . ack 1 win 4096
789 rtsg.1023 > csam.login: P 1:2(1) ack 1 win 4096
790 csam.login > rtsg.1023: . ack 2 win 4096
791 rtsg.1023 > csam.login: P 2:21(19) ack 1 win 4096
792 csam.login > rtsg.1023: P 1:2(1) ack 21 win 4077
793 csam.login > rtsg.1023: P 2:3(1) ack 21 win 4077 urg 1
794 csam.login > rtsg.1023: P 3:4(1) ack 21 win 4077 urg 1\fP\s+2
798 The first line says that tcp port 1023 on rtsg sent a packet
800 on csam. The \fBS\fP indicates that the \fISYN\fP flag was set.
801 The packet sequence number was 768512 and it contained no data.
802 (The notation is `first:last(nbytes)' which means `sequence
804 up to but not including \fIlast\fP which is \fInbytes\fP bytes of user data'.)
805 There was no piggy-backed ack, the available receive window was 4096
806 bytes and there was a max-segment-size option requesting an mss of
809 Csam replies with a similar packet except it includes a piggy-backed
810 ack for rtsg's SYN. Rtsg then acks csam's SYN. The `.' means no
812 The packet contained no data so there is no data sequence number.
813 Note that the ack sequence
814 number is a small integer (1). The first time \fBtcpdump\fP sees a
815 tcp `conversation', it prints the sequence number from the packet.
816 On subsequent packets of the conversation, the difference between
817 the current packet's sequence number and this initial sequence number
818 is printed. This means that sequence numbers after the
819 first can be interpreted
820 as relative byte positions in the conversation's data stream (with the
821 first data byte each direction being `1'). `-S' will override this
822 feature, causing the original sequence numbers to be output.
824 On the 6th line, rtsg sends csam 19 bytes of data (bytes 2 through 20
825 in the rtsg \(-> csam side of the conversation).
826 The PUSH flag is set in the packet.
827 On the 7th line, csam says it's received data sent by rtsg up to
828 but not including byte 21. Most of this data is apparently sitting in the
829 socket buffer since csam's receive window has gotten 19 bytes smaller.
830 Csam also sends one byte of data to rtsg in this packet.
831 On the 8th and 9th lines,
832 csam sends two bytes of urgent, pushed data to rtsg.
834 If the snapshot was small enough that \fBtcpdump\fP didn't capture
835 the full TCP header, it interprets as much of the header as it can
836 and then reports ``[|\fItcp\fP]'' to indicate the remainder could not
837 be interpreted. If the header contains a bogus option (one with a length
838 that's either too small or beyond the end of the header), tcpdump reports
839 it as ``[\fIbad opt\fP]'' and does not interpret any further options (since
840 it's impossible to tell where they start). If the header length indicates
841 options are present but the IP datagram length is not long enough for the
842 options to actually be there, tcpdump reports it as ``[\fIbad hdr length\fP]''.
847 UDP format is illustrated by this rwho packet:
851 \f(CWactinide.who > broadcast.who: udp 84\fP
855 This says that port \fIwho\fP on host \fIactinide\fP sent a udp
856 datagram to port \fIwho\fP on host \fIbroadcast\fP, the Internet
857 broadcast address. The packet contained 84 bytes of user data.
859 Some UDP services are recognized (from the source or destination
860 port number) and the higher level protocol information printed.
861 In particular, Domain Name service requests (RFC-1034/1035) and Sun
862 RPC calls (RFC-1050) to NFS.
864 UDP Name Server Requests
866 \fI(N.B.:The following description assumes familiarity with
867 the Domain Service protocol described in RFC-1035. If you are not familiar
868 with the protocol, the following description will appear to be written
871 Name server requests are formatted as
875 \fIsrc > dst: id op? flags qtype qclass name (len)\fP
877 \f(CWh2opolo.1538 > helios.domain: 3+ A? ucbvax.berkeley.edu. (37)\fP
881 Host \fIh2opolo\fP asked the domain server on \fIhelios\fP for an
882 address record (qtype=A) associated with the name \fIucbvax.berkeley.edu.\fP
883 The query id was `3'. The `+' indicates the \fIrecursion desired\fP flag
884 was set. The query length was 37 bytes, not including the UDP and
885 IP protocol headers. The query operation was the normal one, \fIQuery\fP,
886 so the op field was omitted. If the op had been anything else, it would
887 have been printed between the `3' and the `+'.
888 Similarly, the qclass was the normal one,
889 \fIC_IN\fP, and omitted. Any other qclass would have been printed
890 immediately after the `A'.
892 A few anomalies are checked and may result in extra fields enclosed in
893 square brackets: If a query contains an answer, name server or
899 are printed as `[\fIn\fPa]', `[\fIn\fPn]' or `[\fIn\fPau]' where \fIn\fP
900 is the appropriate count.
901 If any of the response bits are set (AA, RA or rcode) or any of the
902 `must be zero' bits are set in bytes two and three, `[b2&3=\fIx\fP]'
903 is printed, where \fIx\fP is the hex value of header bytes two and three.
905 UDP Name Server Responses
907 Name server responses are formatted as
911 \fIsrc > dst: id op rcode flags a/n/au type class data (len)\fP
913 \f(CWhelios.domain > h2opolo.1538: 3 3/3/7 A 128.32.137.3 (273)
914 helios.domain > h2opolo.1537: 2 NXDomain* 0/1/0 (97)\fP
918 In the first example, \fIhelios\fP responds to query id 3 from \fIh2opolo\fP
919 with 3 answer records, 3 name server records and 7 authority records.
920 The first answer record is type A (address) and its data is internet
921 address 128.32.137.3. The total size of the response was 273 bytes,
922 excluding UDP and IP headers. The op (Query) and response code
923 (NoError) were omitted, as was the class (C_IN) of the A record.
925 In the second example, \fIhelios\fP responds to query 2 with a
926 response code of non-existent domain (NXDomain) with no answers,
927 one name server and no authority records. The `*' indicates that
928 the \fIauthoritative answer\fP bit was set. Since there were no
929 answers, no type, class or data were printed.
931 Other flag characters that might appear are `\-' (recursion available,
932 RA, \fInot\fP set) and `|' (truncated message, TC, set). If the
933 `question' section doesn't contain exactly one entry, `[\fIn\fPq]'
936 Note that name server requests and responses tend to be large and the
937 default \fIsnaplen\fP of 68 bytes may not capture enough of the packet
938 to print. Use the \fB\-s\fP flag to increase the snaplen if you
939 need to seriously investigate name server traffic. `\fB\-s 128\fP'
940 has worked well for me.
943 NFS Requests and Replies
945 Sun NFS (Network File System) requests and replies are printed as:
949 \fIsrc.xid > dst.nfs: len op args\fP
950 \fIsrc.nfs > dst.xid: reply stat len op results\fP
953 sushi.6709 > wrl.nfs: 112 readlink fh 21,24/10.73165
954 wrl.nfs > sushi.6709: reply ok 40 readlink "../var"
955 sushi.201b > wrl.nfs:
956 144 lookup fh 9,74/4096.6878 "xcolors"
957 wrl.nfs > sushi.201b:
958 reply ok 128 lookup fh 9,74/4134.3150
963 In the first line, host \fIsushi\fP sends a transaction with id \fI6709\fP
964 to \fIwrl\fP (note that the number following the src host is a
965 transaction id, \fInot\fP the source port). The request was 112 bytes,
966 excluding the UDP and IP headers. The operation was a \fIreadlink\fP
967 (read symbolic link) on file handle (\fIfh\fP) 21,24/10.731657119.
968 (If one is lucky, as in this case, the file handle can be interpreted
969 as a major,minor device number pair, followed by the inode number and
971 \fIWrl\fP replies `ok' with the contents of the link.
973 In the third line, \fIsushi\fP asks \fIwrl\fP to lookup the name
974 `\fIxcolors\fP' in directory file 9,74/4096.6878. Note that the data printed
975 depends on the operation type. The format is intended to be self
976 explanatory if read in conjunction with
977 an NFS protocol spec.
979 If the \-v (verbose) flag is given, additional information is printed.
985 sushi.1372a > wrl.nfs:
986 148 read fh 21,11/12.195 8192 bytes @ 24576
987 wrl.nfs > sushi.1372a:
988 reply ok 1472 read REG 100664 ids 417/0 sz 29388
993 (\-v also prints the IP header TTL, ID, and fragmentation fields,
994 which have been omitted from this example.) In the first line,
995 \fIsushi\fP asks \fIwrl\fP to read 8192 bytes from file 21,11/12.195,
996 at byte offset 24576. \fIWrl\fP replies `ok'; the packet shown on the
997 second line is the first fragment of the reply, and hence is only 1472
998 bytes long (the other bytes will follow in subsequent fragments, but
999 these fragments do not have NFS or even UDP headers and so might not be
1000 printed, depending on the filter expression used). Because the \-v flag
1001 is given, some of the file attributes (which are returned in addition
1002 to the file data) are printed: the file type (``REG'', for regular file),
1003 the file mode (in octal), the uid and gid, and the file size.
1005 If the \-v flag is given more than once, even more details are printed.
1007 Note that NFS requests are very large and much of the detail won't be printed
1008 unless \fIsnaplen\fP is increased. Try using `\fB\-s 192\fP' to watch
1011 NFS reply packets do not explicitly identify the RPC operation. Instead,
1012 \fItcpdump\fP keeps track of ``recent'' requests, and matches them to the
1013 replies using the transaction ID. If a reply does not closely follow the
1014 corresponding request, it might not be parsable.
1016 KIP Appletalk (DDP in UDP)
1018 Appletalk DDP packets encapsulated in UDP datagrams are de-encapsulated
1019 and dumped as DDP packets (i.e., all the UDP header information is
1020 discarded). The file
1022 is used to translate appletalk net and node numbers to names.
1023 Lines in this file have the form
1035 The first two lines give the names of appletalk networks. The third
1036 line gives the name of a particular host (a host is distinguished
1037 from a net by the 3rd octet in the number \-
1038 a net number \fImust\fP have two octets and a host number \fImust\fP
1039 have three octets.) The number and name should be separated by
1040 whitespace (blanks or tabs).
1043 file may contain blank lines or comment lines (lines starting with
1046 Appletalk addresses are printed in the form
1052 \f(CW144.1.209.2 > icsd-net.112.220
1053 office.2 > icsd-net.112.220
1054 jssmag.149.235 > icsd-net.2\fP
1060 doesn't exist or doesn't contain an entry for some appletalk
1061 host/net number, addresses are printed in numeric form.)
1062 In the first example, NBP (DDP port 2) on net 144.1 node 209
1063 is sending to whatever is listening on port 220 of net icsd node 112.
1064 The second line is the same except the full name of the source node
1065 is known (`office'). The third line is a send from port 235 on
1066 net jssmag node 149 to broadcast on the icsd-net NBP port (note that
1067 the broadcast address (255) is indicated by a net name with no host
1068 number \- for this reason it's a good idea to keep node names and
1069 net names distinct in /etc/atalk.names).
1071 NBP (name binding protocol) and ATP (Appletalk transaction protocol)
1072 packets have their contents interpreted. Other protocols just dump
1073 the protocol name (or number if no name is registered for the
1074 protocol) and packet size.
1076 \fBNBP packets\fP are formatted like the following examples:
1080 \s-2\f(CWicsd-net.112.220 > jssmag.2: nbp-lkup 190: "=:LaserWriter@*"
1081 jssmag.209.2 > icsd-net.112.220: nbp-reply 190: "RM1140:LaserWriter@*" 250
1082 techpit.2 > icsd-net.112.220: nbp-reply 190: "techpit:LaserWriter@*" 186\fP\s+2
1086 The first line is a name lookup request for laserwriters sent by net icsd host
1087 112 and broadcast on net jssmag. The nbp id for the lookup is 190.
1088 The second line shows a reply for this request (note that it has the
1089 same id) from host jssmag.209 saying that it has a laserwriter
1090 resource named "RM1140" registered on port 250. The third line is
1091 another reply to the same request saying host techpit has laserwriter
1092 "techpit" registered on port 186.
1094 \fBATP packet\fP formatting is demonstrated by the following example:
1098 \s-2\f(CWjssmag.209.165 > helios.132: atp-req 12266<0-7> 0xae030001
1099 helios.132 > jssmag.209.165: atp-resp 12266:0 (512) 0xae040000
1100 helios.132 > jssmag.209.165: atp-resp 12266:1 (512) 0xae040000
1101 helios.132 > jssmag.209.165: atp-resp 12266:2 (512) 0xae040000
1102 helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
1103 helios.132 > jssmag.209.165: atp-resp 12266:4 (512) 0xae040000
1104 helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
1105 helios.132 > jssmag.209.165: atp-resp 12266:6 (512) 0xae040000
1106 helios.132 > jssmag.209.165: atp-resp*12266:7 (512) 0xae040000
1107 jssmag.209.165 > helios.132: atp-req 12266<3,5> 0xae030001
1108 helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
1109 helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
1110 jssmag.209.165 > helios.132: atp-rel 12266<0-7> 0xae030001
1111 jssmag.209.133 > helios.132: atp-req* 12267<0-7> 0xae030002\fP\s+2
1115 Jssmag.209 initiates transaction id 12266 with host helios by requesting
1116 up to 8 packets (the `<0-7>'). The hex number at the end of the line
1117 is the value of the `userdata' field in the request.
1119 Helios responds with 8 512-byte packets. The `:digit' following the
1120 transaction id gives the packet sequence number in the transaction
1121 and the number in parens is the amount of data in the packet,
1122 excluding the atp header. The `*' on packet 7 indicates that the
1125 Jssmag.209 then requests that packets 3 & 5 be retransmitted. Helios
1126 resends them then jssmag.209 releases the transaction. Finally,
1127 jssmag.209 initiates the next request. The `*' on the request
1128 indicates that XO (`exactly once') was \fInot\fP set.
1133 Fragmented Internet datagrams are printed as
1137 \fB(frag \fIid\fB:\fIsize\fB@\fIoffset\fB+)\fR
1138 \fB(frag \fIid\fB:\fIsize\fB@\fIoffset\fB)\fR
1142 (The first form indicates there are more fragments. The second
1143 indicates this is the last fragment.)
1145 \fIId\fP is the fragment id. \fISize\fP is the fragment
1146 size (in bytes) excluding the IP header. \fIOffset\fP is this
1147 fragment's offset (in bytes) in the original datagram.
1149 The fragment information is output for each fragment. The first
1150 fragment contains the higher level protocol header and the frag
1151 info is printed after the protocol info. Fragments
1152 after the first contain no higher level protocol header and the
1153 frag info is printed after the source and destination addresses.
1154 For example, here is part of an ftp from arizona.edu to lbl-rtsg.arpa
1155 over a CSNET connection that doesn't appear to handle 576 byte datagrams:
1159 \s-2\f(CWarizona.ftp-data > rtsg.1170: . 1024:1332(308) ack 1 win 4096 (frag 595a:328@0+)
1160 arizona > rtsg: (frag 595a:204@328)
1161 rtsg.1170 > arizona.ftp-data: . ack 1536 win 2560\fP\s+2
1165 There are a couple of things to note here: First, addresses in the
1166 2nd line don't include port numbers. This is because the TCP
1167 protocol information is all in the first fragment and we have no idea
1168 what the port or sequence numbers are when we print the later fragments.
1169 Second, the tcp sequence information in the first line is printed as if there
1170 were 308 bytes of user data when, in fact, there are 512 bytes (308 in
1171 the first frag and 204 in the second). If you are looking for holes
1172 in the sequence space or trying to match up acks
1173 with packets, this can fool you.
1175 A packet with the IP \fIdon't fragment\fP flag is marked with a
1176 trailing \fB(DF)\fP.
1180 By default, all output lines are preceded by a timestamp. The timestamp
1181 is the current clock time in the form
1187 and is as accurate as the kernel's clock.
1188 The timestamp reflects the time the kernel first saw the packet. No attempt
1189 is made to account for the time lag between when the
1190 ethernet interface removed the packet from the wire and when the kernel
1191 serviced the `new packet' interrupt.
1197 Steven McCanne, all of the
1198 Lawrence Berkeley National Laboratory, University of California, Berkeley, CA.
1200 The current version is available via anonymous ftp:
1203 .I ftp://ftp.ee.lbl.gov/tcpdump.tar.Z
1206 Please send bug reports to tcpdump@ee.lbl.gov.
1208 NIT doesn't let you watch your own outbound traffic, BPF will.
1209 We recommend that you use the latter.
1211 Some attempt should be made to reassemble IP fragments or, at least
1212 to compute the right length for the higher level protocol.
1214 Name server inverse queries are not dumped correctly: The (empty)
1215 question section is printed rather than real query in the answer
1216 section. Some believe that inverse queries are themselves a bug and
1217 prefer to fix the program generating them rather than tcpdump.
1219 Apple Ethertalk DDP packets could be dumped as easily as KIP DDP
1221 Even if we were inclined to do anything to promote the use of
1222 Ethertalk (we aren't), LBL doesn't allow Ethertalk on any of its
1223 networks so we'd would have no way of testing this code.
1225 A packet trace that crosses a daylight savings time change will give
1226 skewed time stamps (the time change is ignored).
1228 Filters expressions that manipulate FDDI headers assume that all FDDI
1229 packets are encapsulated Ethernet packets. This is true for IP, ARP,
1230 and DECNET Phase IV, but is not true for protocols such as ISO CLNS.
1231 Therefore, the filter may inadvertently accept certain packets that
1232 do not properly match the filter expression.