xref: /openbsd/usr.sbin/traceroute/traceroute.8 (revision 4bdff4be)
1.\"	$OpenBSD: traceroute.8,v 1.75 2022/12/23 07:16:55 jmc Exp $
2.\"	$NetBSD: traceroute.8,v 1.6 1995/10/12 03:05:50 mycroft Exp $
3.\"
4.\" Copyright (c) 1990, 1991, 1993
5.\"	The Regents of the University of California.  All rights reserved.
6.\"
7.\" This code is derived from software contributed to Berkeley by
8.\" Van Jacobson.
9.\"
10.\" Redistribution and use in source and binary forms, with or without
11.\" modification, are permitted provided that the following conditions
12.\" are met:
13.\" 1. Redistributions of source code must retain the above copyright
14.\"    notice, this list of conditions and the following disclaimer.
15.\" 2. Redistributions in binary form must reproduce the above copyright
16.\"    notice, this list of conditions and the following disclaimer in the
17.\"    documentation and/or other materials provided with the distribution.
18.\" 3. Neither the name of the University nor the names of its contributors
19.\"    may be used to endorse or promote products derived from this software
20.\"    without specific prior written permission.
21.\"
22.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32.\" SUCH DAMAGE.
33.\"
34.\"	@(#)traceroute.8	8.1 (Berkeley) 6/6/93
35.\"
36.Dd $Mdocdate: December 23 2022 $
37.Dt TRACEROUTE 8
38.Os
39.Sh NAME
40.Nm traceroute ,
41.Nm traceroute6
42.Nd print the route packets take to network host
43.Sh SYNOPSIS
44.Nm traceroute\ \&
45.Op Fl ADdIlnSvx
46.Op Fl f Ar first_ttl
47.Op Fl g Ar gateway_addr
48.Op Fl m Ar max_ttl
49.Op Fl P Ar proto
50.Op Fl p Ar port
51.Op Fl q Ar nqueries
52.Op Fl s Ar sourceaddr
53.Op Fl t Ar toskeyword
54.Op Fl V Ar rtable
55.Op Fl w Ar waittime
56.Ar host
57.Op Ar datalen
58.Nm traceroute6
59.Op Fl ADdIlnSv
60.Op Fl f Ar first_hop
61.Op Fl m Ar max_hop
62.Op Fl p Ar port
63.Op Fl q Ar nqueries
64.Op Fl s Ar sourceaddr
65.Op Fl t Ar toskeyword
66.Op Fl V Ar rtable
67.Op Fl w Ar waittime
68.Ar host
69.Op Ar datalen
70.Sh DESCRIPTION
71The Internet is a large and complex aggregation of
72network hardware, connected together by gateways.
73Tracking the route packets follow (or finding the miscreant
74gateway that's discarding packets) can be difficult.
75.Nm
76and
77.Nm traceroute6
78attempt to elicit
79.Dv TIME_EXCEEDED
80responses from each gateway along the path to
81.Ar host ,
82in order to determine their route.
83.Nm
84is used for IPv4 networks and
85.Nm traceroute6
86for IPv6.
87.Pp
88The options are as follows:
89.Bl -tag -width Ds
90.It Fl A
91Look up the AS number for each hop address.
92Uses the DNS service described at
93.Lk https://www.team-cymru.com/IP-ASN-mapping.html#dns
94.It Fl D
95Dump the packet data to standard error before transmitting it.
96.It Fl d
97Turn on socket-level debugging.
98.It Fl f Ar first_ttl
99Set the first TTL or hop limit used in outgoing probe packets.
100The effect is that the first first_ttl \- 1 hosts will be skipped
101in the output of
102.Nm traceroute .
103The default is 1 (skip no hosts).
104.It Fl g Ar gateway_addr
105Add
106.Ar gateway_addr
107to the list of addresses in the IP Loose Source Record Route (LSRR)
108option.
109If no gateways are specified, the LSRR option is omitted.
110This option is not available for IPv6.
111.It Fl I
112Use ICMP or ICMP6 ECHO instead of UDP datagrams.
113.It Fl l
114Display the TTL or hop limit value of the returned packet.
115This is useful for checking for asymmetric routing.
116.It Fl m Ar max_ttl
117Set the maximum TTL or hop limit.
118The default is the value of the system's
119.Va net.inet.ip.ttl
120or
121.Va net.inet6.ip6.hlim
122.Xr sysctl 8
123variable, which defaults to 64.
124.It Fl n
125Print hop addresses numerically rather than symbolically and numerically
126(saves a nameserver address-to-name lookup for each gateway found on the
127path).
128.It Fl P Ar proto
129Change the protocol being used from UDP
130to a numeric protocol or a name as specified in
131.Pa /etc/protocols .
132This will not work reliably for most protocols.
133If set to 1 (ICMP), then
134ICMP Echo Request messages will be used (same as
135.Xr ping 8 ) .
136This option is not available for IPv6.
137.It Fl p Ar port
138Set the base UDP
139.Ar port
140number used in probes.
141The default is 33434.
142.Nm
143hopes that nothing is listening on UDP ports
144.Ar base
145to
146.Ar base Ns + Ns Ar nhops Ns * Ns Ar nqueries Ns -1
147at the destination host (so an ICMP
148.Dv PORT_UNREACHABLE
149message will
150be returned to terminate the route tracing).
151If something is
152listening on a port in the default range, this option can be used
153to pick an unused port range.
154.It Fl q Ar nqueries
155Set the number of probes per TTL to
156.Ar nqueries .
157The default is three probes.
158.It Fl S
159Print how many probes were not answered for each hop.
160.It Fl s Ar sourceaddr
161Set the source address to transmit from, which is useful on machines
162with multiple interfaces.
163.It Fl t Ar toskeyword
164Set the type-of-service (TOS) in probe packets.
165The value may be one of
166.Cm critical ,
167.Cm inetcontrol ,
168.Cm lowdelay ,
169.Cm netcontrol ,
170.Cm throughput ,
171.Cm reliability ,
172or one of the DiffServ Code Points:
173.Cm ef ,
174.Cm af11 ... af43 ,
175.Cm cs0 ... cs7 ;
176or a number in either hex or decimal.
177The default is zero.
178This option can be used to
179see if different types-of-service result in different paths.
180If this option is used, changes to the type-of-service in the
181returned packets are displayed.
182Not all values of TOS are legal or meaningful \-
183see the IP spec for definitions.
184Useful values are probably
185.Cm lowdelay
186and
187.Cm throughput .
188.It Fl V Ar rtable
189Set the routing table to be used.
190.It Fl v
191Verbose output.
192Received ICMP packets other than
193.Dv TIME_EXCEEDED
194and
195.Dv UNREACHABLE Ns s
196are listed.
197.It Fl w Ar waittime
198Set the time, in seconds, to wait for a response to a probe.
199The default is 3.
200.It Fl x
201Print the ICMP extended headers if available.
202This option is not available for IPv6.
203.It Ar host
204The destination host,
205specified as a host name or numeric IP address.
206.It Ar datalen
207The probe datagram length.
208The default is 40 bytes for IPv4 UDP
209and 60 bytes for ICMP, IPv6 UDP and ICMP6.
210.El
211.Pp
212The program attempts to trace the route an IP packet would follow to a
213host by launching UDP probe packets with a small TTL or hop limit,
214then listening for an ICMP "time exceeded" reply from a gateway.
215It starts using probes with a TTL/hop limit of one
216and increases by one until it gets an ICMP "port unreachable"
217(which means it reached the host) or hits a maximum limit
218(which defaults to 64, but can be changed using the
219.Fl m
220option).
221Three probes (the exact number can be changed using the
222.Fl q
223option) are sent and a line is printed
224showing the TTL or hop limit, address of the gateway,
225and round trip time of each probe.
226If the probe answers come from different gateways,
227the address of each responding system will be printed.
228If there is no response within a 3 second timeout
229interval (which can be changed using the
230.Fl w
231option), a "*" is printed for that
232probe.
233If the host cannot be reached,
234.Nm
235skips printing lines consisting only of * until the maximum TTL/hop limit is
236reached.
237.Pp
238We don't want the destination
239host to process the UDP
240probe packets so the destination port is set to an
241unlikely value (if some clod on the destination is using that
242value, it can be changed using the
243.Fl p
244option).
245.Pp
246A sample use and output might be:
247.Bd -literal -offset indent
248$ traceroute nis.nsf.net.
249traceroute to nis.nsf.net (35.1.1.48), 64 hops max, 56 byte packet
2501  helios.ee.lbl.gov (128.3.112.1)  19 ms  19 ms  0 ms
2512  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  39 ms  19 ms
2523  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  39 ms  19 ms
2534  ccngw-ner-cc.Berkeley.EDU (128.32.136.23)  39 ms  40 ms  39 ms
2545  ccn-nerif22.Berkeley.EDU (128.32.168.22)  39 ms  39 ms  39 ms
2556  128.32.197.4 (128.32.197.4)  40 ms  59 ms  59 ms
2567  131.119.2.5 (131.119.2.5)  59 ms  59 ms  59 ms
2578  129.140.70.13 (129.140.70.13)  99 ms  99 ms  80 ms
2589  129.140.71.6 (129.140.71.6)  139 ms  239 ms  319 ms
25910  129.140.81.7 (129.140.81.7)  220 ms  199 ms  199 ms
26011  nic.merit.edu (35.1.1.48)  239 ms  239 ms  239 ms
261.Ed
262.Pp
263Note that lines 2 & 3 are the same.
264This is due to a buggy
265kernel on the 2nd hop system \- lbl-csam.arpa \- that forwards
266packets with a zero TTL (a bug in the distributed version of
267.Bx 4.3 ) .
268Note that you have to guess what path
269the packets are taking cross-country since the NSFNET (129.140)
270doesn't supply address-to-name translations for its NSSes.
271.Pp
272A more interesting example is:
273.Bd -literal -offset indent
274$ traceroute allspice.lcs.mit.edu.
275traceroute to allspice.lcs.mit.edu (18.26.0.115), 64 hops max
2761  helios.ee.lbl.gov (128.3.112.1)  0 ms  0 ms  0 ms
2772  lilac-dmc.Berkeley.EDU (128.32.216.1)  19 ms  19 ms  19 ms
2783  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  19 ms  19 ms
2794  ccngw-ner-cc.Berkeley.EDU (128.32.136.23)  19 ms  39 ms  39 ms
2805  ccn-nerif22.Berkeley.EDU (128.32.168.22)  20 ms  39 ms  39 ms
2816  128.32.197.4 (128.32.197.4)  59 ms  119 ms  39 ms
2827  131.119.2.5 (131.119.2.5)  59 ms  59 ms  39 ms
2838  129.140.70.13 (129.140.70.13)  80 ms  79 ms  99 ms
2849  129.140.71.6 (129.140.71.6)  139 ms  139 ms  159 ms
28510  129.140.81.7 (129.140.81.7)  199 ms  180 ms  300 ms
28611  129.140.72.17 (129.140.72.17)  300 ms  239 ms  239 ms
28712  * * *
28813  128.121.54.72 (128.121.54.72)  259 ms  499 ms  279 ms
28914  * * *
29015  * * *
29116  * * *
29217  * * *
29318  ALLSPICE.LCS.MIT.EDU (18.26.0.115)  339 ms  279 ms  279 ms
294.Ed
295.Pp
296Note that the gateways 12, 14, 15, 16 & 17 hops away
297either don't send ICMP "time exceeded" messages or send them
298with a TTL too small to reach us.
29914 \- 17 are running the MIT
300C Gateway code that doesn't send "time exceeded"s.
301God only knows what's going on with 12.
302.Pp
303The silent gateway 12 in the above may be the result of a bug in
304the 4.[23]
305.Bx
306network code (and its derivatives):  4.x (x <= 3)
307sends an unreachable message using whatever TTL remains in the
308original datagram.
309Since, for gateways, the remaining TTL is zero, the ICMP
310"time exceeded" is guaranteed to not make it back to us.
311The behavior of this bug is slightly more interesting
312when it appears on the destination system:
313.Bd -literal -offset indent
3141  helios.ee.lbl.gov (128.3.112.1)  0 ms  0 ms  0 ms
3152  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  19 ms  39 ms
3163  lilac-dmc.Berkeley.EDU (128.32.216.1)  19 ms  39 ms  19 ms
3174  ccngw-ner-cc.Berkeley.EDU (128.32.136.23)  39 ms  40 ms  19 ms
3185  ccn-nerif35.Berkeley.EDU (128.32.168.35)  39 ms  39 ms  39 ms
3196  csgw.Berkeley.EDU (128.32.133.254)  39 ms  59 ms  39 ms
3207  * * *
3218  * * *
3229  * * *
32310  * * *
32411  * * *
32512  * * *
32613  rip.Berkeley.EDU (128.32.131.22)  59 ms !  39 ms !  39 ms !
327.Ed
328.Pp
329Notice that there are 12 "gateways" (13 is the final
330destination) and exactly the last half of them are "missing".
331What's really happening is that rip (a Sun-3 running Sun OS3.5)
332is using the TTL from our arriving datagram as the TTL in its
333ICMP reply.
334So, the reply will time out on the return path
335(with no notice sent to anyone since ICMPs aren't sent for ICMPs)
336until we probe with a TTL that's at least twice the path
337length.
338That is, rip is really only 7 hops away.
339A reply that returns with a TTL of 1 is a clue this problem exists.
340.Nm
341prints a "!" after the time if the TTL is <= 1.
342Since vendors ship a lot of obsolete (DEC's Ultrix, Sun 3.x) or
343non-standard (HP-UX) software, expect to see this problem
344frequently and/or take care picking the target host of your
345probes.
346.Pp
347Other possible annotations after the time are
348.Sy !H ,
349.Sy !N ,
350.Sy !P
351(got a host, network or protocol unreachable, respectively),
352.Sy !A ,
353.Sy !C
354(access to the network or host, respectively, is prohibited),
355.Sy !X
356(communication administratively prohibited by filtering),
357.Sy !S
358or
359.Sy !F
360(source route failed or fragmentation needed \- neither of these should
361ever occur and the associated gateway is busted if you see one),
362.Sy !U
363(destination network or host unknown),
364.Sy !T
365(destination network or host unreachable for TOS),
366.Sy !<code>
367(other ICMP unreachable code).
368.Sy TOS=xxx!
369(TOS bit in returned packet differs from last hop).
370If almost all the probes result in some kind of unreachable,
371.Nm
372will give up and exit.
373.Pp
374.Dl $ traceroute -g 10.3.0.5 128.182.0.0
375.Pp
376will show the path from the Cambridge Mailbridge to PSC, while
377.Pp
378.Dl $ traceroute -g 192.5.146.4 -g 10.3.0.5 35.0.0.0
379.Pp
380will show the path from the Cambridge Mailbridge to Merit, using PSC to
381reach the Mailbridge.
382.Pp
383This program is intended for use in network testing, measurement
384and management.
385It should be used primarily for manual fault isolation.
386Because of the load it could impose on the network, it is unwise to use
387.Nm
388during normal operations or from automated scripts.
389.Sh SEE ALSO
390.Xr ping 8 ,
391.Xr route 8
392.Sh HISTORY
393The very first
394.Nm
395(never released) used ICMP ECHO_REQUEST
396datagrams as probe packets.
397During the first night of testing it was
398discovered that more than half the router vendors of the time would
399not return an ICMP TIME_EXCEEDED for an ECHO_REQUEST.
400.Nm
401was then changed to use UDP probe packets.
402Most modern TCP/IP implementations will now generate an ICMP error
403message to ICMP query messages, and the option to use ECHO_REQUEST probes
404was re-implemented.
405.Pp
406The
407.Nm
408command first appeared in
409.Bx 4.3 Reno .
410The
411.Nm traceroute6
412command first appeared in the WIDE Hydrangea IPv6 protocol stack kit.
413.Sh AUTHORS
414.An -nosplit
415Implemented by
416.An Van Jacobson
417from a suggestion by
418.An Steve Deering .
419Debugged
420by a cast of thousands with particularly cogent suggestions or fixes from
421.An C. Philip Wood ,
422.An Tim Seaver ,
423and
424.An Ken Adelman .
425