xref: /freebsd/usr.sbin/traceroute/traceroute.8 (revision 1edb7116)
1.\" Copyright (c) 1989, 1995, 1996, 1997, 1999, 2000
2.\"	The Regents of the University of California.  All rights reserved.
3.\"
4.\" Redistribution and use in source and binary forms are permitted
5.\" provided that the above copyright notice and this paragraph are
6.\" duplicated in all such forms and that any documentation,
7.\" advertising materials, and other materials related to such
8.\" distribution and use acknowledge that the software was developed
9.\" by the University of California, Berkeley.  The name of the
10.\" University may not be used to endorse or promote products derived
11.\" from this software without specific prior written permission.
12.\" THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR
13.\" IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
14.\" WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE.
15.\"
16.\"	$Id: traceroute.8,v 1.19 2000/09/21 08:44:19 leres Exp $
17.\"
18.Dd November 17, 2023
19.Dt TRACEROUTE 8
20.Os
21.Sh NAME
22.Nm traceroute
23.Nd "print the route packets take to network host"
24.Sh SYNOPSIS
25.Nm
26.Bk -words
27.Op Fl adDeEFISnrvx
28.Op Fl A Ar as_server
29.Op Fl f Ar first_ttl
30.Op Fl g Ar gateway
31.Op Fl i Ar iface
32.Op Fl m Ar max_ttl
33.Op Fl M Ar first_ttl
34.Op Fl p Ar port
35.Op Fl P Ar proto
36.Op Fl q Ar nprobes
37.Op Fl s Ar src_addr
38.Op Fl t Ar tos
39.Op Fl w Ar waittime
40.Op Fl z Ar pausemsecs
41.Ar host
42.Op Ar packetlen
43.Ek
44.Sh DESCRIPTION
45The Internet is a large and complex aggregation of network hardware, connected
46together by gateways.
47Tracking the route one's packets follow (or finding the miscreant gateway
48that's discarding your packets) can be difficult.
49.Nm
50utilizes the IP protocol `time to live' field and attempts to elicit an ICMP
51TIME_EXCEEDED response from each gateway along the path to some host.
52.Pp
53The only mandatory parameter is the destination host name or IP number.
54The default probe datagram length is 40 bytes, but this may be increased by
55specifying a packet length (in bytes) after the destination host name.
56.Pp
57Other options are:
58.Bl -tag -width Ds
59.It Fl a
60Turn on AS# lookups for each hop encountered.
61.It Fl A Ar as_server
62Turn on AS# lookups and use the given server instead of the default.
63.It Fl d
64Enable socket level debugging.
65.It Fl D
66When an ICMP response to our probe datagram is received, print the differences
67between the transmitted packet and the packet quoted by the ICMP response.
68A key showing the location of fields within the transmitted packet is printed,
69followed by the original packet in hex, followed by the quoted packet in hex.
70Bytes that are unchanged in the quoted packet are shown as underscores.
71Note, the IP checksum and the TTL of the quoted packet are not expected to
72match.
73By default, only one probe per hop is sent with this option.
74.It Fl e
75Firewall evasion mode.
76Use fixed destination ports for UDP, UDP-Lite, TCP and SCTP probes.
77The destination port does NOT increment with each packet sent.
78.It Fl E
79Detect ECN bleaching.
80Set the
81.Em IPTOS_ECN_ECT1
82Explicit Congestion Notification (ECN) bits
83.Pq Dv 01 ,
84and report if the hop has bleached
85.Pq Dv 00
86or mangled
87.Pq Dv 10
88them, or if it is experiencing congestion
89.Pq Dv 11 .
90Otherwise, report that it passed the bits appropriately.
91If
92.Fl t
93is also specified, the corresponding ECN bits will be replaced.
94.It Fl f Ar first_ttl
95Set the initial time-to-live used in the first outgoing probe packet.
96.It Fl F
97Set the "don't fragment" bit.
98.It Fl g Ar gateway
99Specify a loose source route gateway (8 maximum).
100.It Fl i Ar iface
101Specify a network interface to obtain the source IP address for outgoing probe
102packets.
103This is normally only useful on a multi-homed host.
104(See the
105.Fl s
106flag for another way to do this).
107.It Fl I
108Use ICMP ECHO instead of UDP datagrams.
109(A synonym for "-P icmp").
110.It Fl m Ar max_ttl
111Set the max time-to-live (max number of hops) used in outgoing probe packets.
112The default is the value of the
113.Va net.inet.ip.ttl
114.Xr sysctl 8
115(the same default used for TCP connections).
116.It Fl M Ar first_ttl
117Set the initial time-to-live value used in outgoing probe packets.
118The default is 1, i.e., start with the first hop.
119.It Fl n
120Print hop addresses numerically rather than symbolically and numerically
121(saves a nameserver address-to-name lookup for each gateway found on the path).
122.It Fl p Ar port
123Protocol specific.
124For UDP, UDP-Lite, TCP and SCTP, sets the base
125.Ar port
126number used in probes (default is 33434).
127Traceroute hopes that nothing is listening on UDP ports (or UDP-Lite ports
128if used by
129.Nm
130and supported by the peer)
131.Em port + 1
132to
133.Em port + (max_ttl - first_ttl + 1) * nprobes
134at the destination host (so an ICMP PORT_UNREACHABLE message will be returned
135to terminate the route tracing).
136If something is listening on a port in the default range, this option can be
137used to pick an unused port range.
138.It Fl P Ar proto
139Send packets of specified IP protocol.
140The currently supported protocols
141are: UDP, UDP-Lite, TCP, SCTP, GRE and ICMP.
142Other protocols may also be specified (either by name or by number), though
143.Nm
144does not implement any special knowledge of their packet formats.
145This option is useful for determining which router along a path may be blocking
146packets based on IP protocol number.
147But see BUGS below.
148.It Fl q Ar nprobes
149Set the number of probes per hop (default is 3, unless
150.Fl D
151is specified,
152when it is 1).
153.It Fl r
154Bypass the normal routing tables and send directly to a host on an attached
155network.
156If the host is not on a directly-attached network, an error is returned.
157This option can be used to ping a local host through an interface that has no
158route through it (e.g., after the interface was dropped by
159.Xr routed 8 .
160.It Fl s Ar src_addr
161Use the following IP address (which usually is given as an IP number, not a
162hostname) as the source address in outgoing probe packets.
163On multi-homed hosts (those with more than one IP address), this option can be
164used to force the source address to be something other than the IP address of
165the interface the probe packet is sent on.
166If the IP address is not one of this machine's interface addresses, an error is
167returned and nothing is sent.
168(See the
169.Fl i
170flag for another way to do this).
171.It Fl S
172Print a summary of how many probes were not answered for each hop.
173.It Fl t Ar tos
174Set the
175.Em type-of-service
176in probe packets to the following value (default zero).
177The value must be a decimal integer in the range 0 to 255.
178This option can be used to see if different types-of-service result in
179different paths.
180The upper six bits are the Differentiated Services Codepoint (RFC4594).
181The lower two bits are the Explicit Congestion Notification field (RFC3168).
182.It Fl v
183Verbose output.
184Received ICMP packets other than
185.Dv TIME_EXCEEDED
186and
187.Dv UNREACHABLE Ns s
188are listed.
189.It Fl w Ar waittime
190Set the time (in seconds) to wait for a response to a probe (default 5 sec.).
191.It Fl x
192Toggle ip checksums.
193Normally, this prevents traceroute from calculating ip checksums.
194In some cases, the operating system can overwrite parts of the outgoing packet
195but not recalculate the checksum (so in some cases the default is to not
196calculate checksums and using
197.Fl x
198causes them to be calculated).
199Note that checksums are usually required for the last hop when using ICMP ECHO
200probes
201.Pq Fl I .
202So they are always calculated when using ICMP.
203.It Fl z Ar pausemsecs
204Set the time (in milliseconds) to pause between probes (default 0).
205Some systems such as Solaris and routers such as Ciscos rate limit ICMP
206messages.
207A good value to use with this is 500 (e.g., 1/2 second).
208.El
209.Pp
210This program attempts to trace the route an IP packet would follow to some
211internet host by launching UDP probe packets with a small TTL (time to live)
212then listening for an ICMP "time exceeded" reply from a gateway.
213We start our probes with a TTL of one and increase by one until we get an ICMP
214"port unreachable" (which means we got to "host") or hit a max (which defaults
215to the amount of hops specified by the
216.Va net.inet.ip.ttl
217.Xr sysctl 8
218and can be changed with the
219.Fl m
220flag).
221Three probes (change with
222.Fl q
223flag) are sent at each TTL setting and a line is printed showing the TTL,
224address of the gateway and round trip time of each probe.
225If the probe answers come from different gateways, the address of each
226responding system will be printed.
227If there is no response within a 5 sec. timeout interval (changed with the
228.Fl w
229flag), a "*" is printed for that probe.
230.Pp
231We don't want the destination host to process the UDP probe packets so the
232destination port is set to an unlikely value (if some clod on the destination
233is using that value, it can be changed with the
234.Fl p
235flag).
236.Pp
237A sample use and output might be:
238.Bd -literal -offset 4n
239% traceroute nis.nsf.net.
240traceroute to nis.nsf.net (35.1.1.48), 64 hops max, 40 byte packets
241 1  helios.ee.lbl.gov (128.3.112.1)  19 ms  19 ms  0 ms
242 2  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  39 ms  19 ms
243 3  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  39 ms  19 ms
244 4  ccngw-ner-cc.Berkeley.EDU (128.32.136.23)  39 ms  40 ms  39 ms
245 5  ccn-nerif22.Berkeley.EDU (128.32.168.22)  39 ms  39 ms  39 ms
246 6  128.32.197.4 (128.32.197.4)  40 ms  59 ms  59 ms
247 7  131.119.2.5 (131.119.2.5)  59 ms  59 ms  59 ms
248 8  129.140.70.13 (129.140.70.13)  99 ms  99 ms  80 ms
249 9  129.140.71.6 (129.140.71.6)  139 ms  239 ms  319 ms
25010  129.140.81.7 (129.140.81.7)  220 ms  199 ms  199 ms
25111  nic.merit.edu (35.1.1.48)  239 ms  239 ms  239 ms
252.Ed
253.Pp
254Note that lines 2 & 3 are the same.
255This is due to a buggy kernel on the 2nd hop system \- lilac-dmc.Berkeley.EDU \-
256that forwards packets with a zero TTL (a bug in the distributed version of
2574.3BSD).
258Note that you have to guess what path the packets are taking cross-country
259since the NSFNet (129.140) doesn't supply address-to-name translations for its
260NSSes.
261.Pp
262A more interesting example is:
263.Bd -literal -offset 4n
264% traceroute allspice.lcs.mit.edu.
265traceroute to allspice.lcs.mit.edu (18.26.0.115), 64 hops max, 40 byte packets
266 1  helios.ee.lbl.gov (128.3.112.1)  0 ms  0 ms  0 ms
267 2  lilac-dmc.Berkeley.EDU (128.32.216.1)  19 ms  19 ms  19 ms
268 3  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  19 ms  19 ms
269 4  ccngw-ner-cc.Berkeley.EDU (128.32.136.23)  19 ms  39 ms  39 ms
270 5  ccn-nerif22.Berkeley.EDU (128.32.168.22)  20 ms  39 ms  39 ms
271 6  128.32.197.4 (128.32.197.4)  59 ms  119 ms  39 ms
272 7  131.119.2.5 (131.119.2.5)  59 ms  59 ms  39 ms
273 8  129.140.70.13 (129.140.70.13)  80 ms  79 ms  99 ms
274 9  129.140.71.6 (129.140.71.6)  139 ms  139 ms  159 ms
27510  129.140.81.7 (129.140.81.7)  199 ms  180 ms  300 ms
27611  129.140.72.17 (129.140.72.17)  300 ms  239 ms  239 ms
27712  * * *
27813  128.121.54.72 (128.121.54.72)  259 ms  499 ms  279 ms
27914  * * *
28015  * * *
28116  * * *
28217  * * *
28318  ALLSPICE.LCS.MIT.EDU (18.26.0.115)  339 ms  279 ms  279 ms
284.Ed
285.Pp
286Note that the gateways 12, 14, 15, 16 & 17 hops away either don't send ICMP
287"time exceeded" messages or send them with a TTL too small to reach us.
28814 \- 17 are running the MIT C Gateway code that doesn't send "time exceeded"s.
289God only knows what's going on with 12.
290.Pp
291The silent gateway 12 in the above may be the result of a bug in the 4.[23]BSD
292network code (and its derivatives):  4.x (x <= 3) sends an unreachable message
293using whatever TTL remains in the original datagram.
294Since, for gateways, the remaining TTL is zero, the ICMP "time exceeded" is
295guaranteed to not make it back to us.
296The behavior of this bug is slightly more interesting when it appears on the
297destination system:
298.Bd -literal -offset 4n
299 1  helios.ee.lbl.gov (128.3.112.1)  0 ms  0 ms  0 ms
300 2  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  19 ms  39 ms
301 3  lilac-dmc.Berkeley.EDU (128.32.216.1)  19 ms  39 ms  19 ms
302 4  ccngw-ner-cc.Berkeley.EDU (128.32.136.23)  39 ms  40 ms  19 ms
303 5  ccn-nerif35.Berkeley.EDU (128.32.168.35)  39 ms  39 ms  39 ms
304 6  csgw.Berkeley.EDU (128.32.133.254)  39 ms  59 ms  39 ms
305 7  * * *
306 8  * * *
307 9  * * *
30810  * * *
30911  * * *
31012  * * *
31113  rip.Berkeley.EDU (128.32.131.22)  59 ms !  39 ms !  39 ms !
312.Ed
313.Pp
314Notice that there are 12 "gateways" (13 is the final destination) and exactly
315the last half of them are "missing".
316What's really happening is that rip (a Sun-3 running Sun OS3.5) is using the
317TTL from our arriving datagram as the TTL in its ICMP reply.
318So, the reply will time out on the return path (with no notice sent to anyone
319since ICMP's aren't sent for ICMP's) until we probe with a TTL that's at least
320twice the path length.
321I.e., rip is really only 7 hops away.
322A reply that returns with a TTL of 1 is a clue this problem exists.
323.Nm
324prints a "!" after the time if the TTL is <= 1.
325Since vendors ship a lot of obsolete
326.Pf ( DEC Ns \'s
327Ultrix, Sun 3.x) or
328non-standard
329.Pq HP-UX
330software, expect to see this problem frequently and/or take care picking the
331target host of your probes.
332.Pp
333Other possible annotations after the time are:
334.Bl -hang -offset indent -width 12n
335.It Sy !H
336Host unreachable.
337.It Sy !N
338Network unreachable.
339.It Sy !P
340Protocol unreachable.
341.It Sy !S
342Source route failed.
343.It Sy !F\-<pmtu>
344Fragmentation needed.
345The RFC1191 Path MTU Discovery value is displayed.
346.It Sy !U
347Destination network unknown.
348.It Sy !W
349Destination host unknown.
350.It Sy !I
351Source host is isolated.
352.It Sy !A
353Communication with destination network administratively prohibited.
354.It Sy !Z
355Communication with destination host administratively prohibited.
356.It Sy !Q
357For this ToS the destination network is unreachable.
358.It Sy !T
359For this ToS the destination host is unreachable.
360.It Sy !X
361Communication administratively prohibited.
362.It Sy !V
363Host precedence violation.
364.It Sy !C
365Precedence cutoff in effect.
366.It Sy !<num>
367ICMP unreachable code <num>.
368.El
369.Pp
370These are defined by RFC1812 (which supersedes RFC1716).
371If almost all the probes result in some kind of unreachable,
372.Nm
373will give up and exit.
374.Pp
375This program is intended for use in network testing, measurement and
376management.
377It should be used primarily for manual fault isolation.
378Because of the load it could impose on the network, it is unwise to use
379.Nm
380during normal operations or from automated scripts.
381.Sh SEE ALSO
382.Xr netstat 1 ,
383.Xr ping 8 ,
384.Xr traceroute6 8
385.Sh AUTHORS
386Implemented by
387.An Van Jacobson
388from a suggestion by Steve Deering.
389Debugged by a cast of thousands with particularly cogent suggestions or fixes
390from C. Philip Wood, Tim Seaver and Ken Adelman.
391.Sh BUGS
392When using protocols other than UDP, functionality is reduced.
393In particular, the last packet will often appear to be lost, because even
394though it reaches the destination host, there's no way to know that because no
395ICMP message is sent back.
396In the TCP case,
397.Nm
398should listen for a RST from the destination host (or an intermediate router
399that's filtering packets), but this is not implemented yet.
400.Pp
401The AS number capability reports information that may sometimes be inaccurate
402due to discrepancies between the contents of the routing database server and
403the current state of the Internet.
404