1PING(8) iputils PING(8)
2
6 ping - send ICMP ECHO_REQUEST to network hosts
7
9 ping [-aAbBdDfhLnOqrRUvV46] [-c count] [-F flowlabel] [-i interval]
10 [-I interface] [-l preload] [-m mark] [-M pmtudisc_option]
11 [-N nodeinfo_option] [-w deadline] [-W timeout] [-p pattern]
12 [-Q tos] [-s packetsize] [-S sndbuf] [-t ttl]
13 [-T timestamp option] [hop...] {destination}
14
16 ping uses the ICMP protocol's mandatory ECHO_REQUEST datagram to elicit
17 an ICMP ECHO_RESPONSE from a host or gateway. ECHO_REQUEST datagrams
18 (“pings”) have an IP and ICMP header, followed by a struct timeval and
19 then an arbitrary number of “pad” bytes used to fill out the packet.
20 ping works with both IPv4 and IPv6. Using only one of them explicitly
22 can be enforced by specifying -4 or -6.
23 ping can also send IPv6 Node Information Queries (RFC4620).
25 Intermediate hops may not be allowed, because IPv6 source routing was
26 deprecated (RFC5095).
27
29 -4
30 Use IPv4 only.
31 -6
33 Use IPv6 only.
34 -a
36 Audible ping.
37 -A
39 Adaptive ping. Interpacket interval adapts to round-trip time, so
40 that effectively not more than one (or more, if preload is set)
41 unanswered probe is present in the network. Minimal interval is
42 200msec unless super-user. On networks with low RTT this mode is
43 essentially equivalent to flood mode.
44 -b
46 Allow pinging a broadcast address.
47 -B
49 Do not allow ping to change source address of probes. The address
50 is bound to one selected when ping starts.
51 -c count
53 Stop after sending count ECHO_REQUEST packets. With deadline
54 option, ping waits for count ECHO_REPLY packets, until the timeout
55 expires.
56 -d
58 Set the SO_DEBUG option on the socket being used. Essentially, this
59 socket option is not used by Linux kernel.
60 -D
62 Print timestamp (unix time + microseconds as in gettimeofday)
63 before each line.
64 -f
66 Flood ping. For every ECHO_REQUEST sent a period “.” is printed,
67 while for every ECHO_REPLY received a backspace is printed. This
68 provides a rapid display of how many packets are being dropped. If
69 interval is not given, it sets interval to zero and outputs packets
70 as fast as they come back or one hundred times per second,
71 whichever is more. Only the super-user may use this option with
72 zero interval.
73 -F flow label
75 IPv6 only. Allocate and set 20 bit flow label (in hex) on echo
76 request packets. If value is zero, kernel allocates random flow
77 label.
78 -h
80 Show help.
81 -i interval
83 Wait interval seconds between sending each packet. Real number
84 allowed with dot as a decimal separator (regardless locale setup).
85 The default is to wait for one second between each packet normally,
86 or not to wait in flood mode. Only super-user may set interval to
87 values less than 0.2 seconds.
88 -I interface
90 interface is either an address, an interface name or a VRF name. If
91 interface is an address, it sets source address to specified
92 interface address. If interface is an interface name, it sets
93 source interface to specified interface. If interface is a VRF
94 name, each packet is routed using the corresponding routing table;
95 in this case, the -I option can be repeated to specify a source
96 address. NOTE: For IPv6, when doing ping to a link-local scope
97 address, link specification (by the '%'-notation in destination, or
98 by this option) can be used but it is no longer required.
99 -l preload
101 If preload is specified, ping sends that many packets not waiting
102 for reply. Only the super-user may select preload more than 3.
103 -L
105 Suppress loopback of multicast packets. This flag only applies if
106 the ping destination is a multicast address.
107 -m mark
109 use mark to tag the packets going out. This is useful for variety
110 of reasons within the kernel such as using policy routing to select
111 specific outbound processing.
112 -M pmtudisc_opt
114 Select Path MTU Discovery strategy. pmtudisc_option may be either
115 do (prohibit fragmentation, even local one), want (do PMTU
116 discovery, fragment locally when packet size is large), or dont (do
117 not set DF flag).
118 -N nodeinfo_option
120 IPv6 only. Send ICMPv6 Node Information Queries (RFC4620), instead
121 of Echo Request. CAP_NET_RAW capability is required.
122 help
124 Show help for NI support.
125 name
127 Queries for Node Names.
128 ipv6
130 Queries for IPv6 Addresses. There are several IPv6 specific
131 flags.
132 ipv6-global
134 Request IPv6 global-scope addresses.
135 ipv6-sitelocal
137 Request IPv6 site-local addresses.
138 ipv6-linklocal
140 Request IPv6 link-local addresses.
141 ipv6-all
143 Request IPv6 addresses on other interfaces.
144 ipv4
146 Queries for IPv4 Addresses. There is one IPv4 specific flag.
147 ipv4-all
149 Request IPv4 addresses on other interfaces.
150 subject-ipv6=ipv6addr
152 IPv6 subject address.
153 subject-ipv4=ipv4addr
155 IPv4 subject address.
156 subject-name=nodename
158 Subject name. If it contains more than one dot, fully-qualified
159 domain name is assumed.
160 subject-fqdn=nodename
162 Subject name. Fully-qualified domain name is always assumed.
163 -n
165 Numeric output only. No attempt will be made to lookup symbolic
166 names for host addresses.
167 -O
169 Report outstanding ICMP ECHO reply before sending next packet. This
170 is useful together with the timestamp -D to log output to a
171 diagnostic file and search for missing answers.
172 -p pattern
174 You may specify up to 16 “pad” bytes to fill out the packet you
175 send. This is useful for diagnosing data-dependent problems in a
176 network. For example, -p ff will cause the sent packet to be filled
177 with all ones.
178 -q
180 Quiet output. Nothing is displayed except the summary lines at
181 startup time and when finished.
182 -Q tos
184 Set Quality of Service -related bits in ICMP datagrams. tos can be
185 decimal (ping only) or hex number.
186 In RFC2474, these fields are interpreted as 8-bit Differentiated
188 Services (DS), consisting of: bits 0-1 (2 lowest bits) of separate
189 data, and bits 2-7 (highest 6 bits) of Differentiated Services
190 Codepoint (DSCP). In RFC2481 and RFC3168, bits 0-1 are used for
191 ECN.
192 Historically (RFC1349, obsoleted by RFC2474), these were
194 interpreted as: bit 0 (lowest bit) for reserved (currently being
195 redefined as congestion control), 1-4 for Type of Service and bits
196 5-7 (highest bits) for Precedence.
197 -r
199 Bypass the normal routing tables and send directly to a host on an
200 attached interface. If the host is not on a directly-attached
201 network, an error is returned. This option can be used to ping a
202 local host through an interface that has no route through it
203 provided the option -I is also used.
204 -R
206 ping only. Record route. Includes the RECORD_ROUTE option in the
207 ECHO_REQUEST packet and displays the route buffer on returned
208 packets. Note that the IP header is only large enough for nine such
209 routes. Many hosts ignore or discard this option.
210 -s packetsize
212 Specifies the number of data bytes to be sent. The default is 56,
213 which translates into 64 ICMP data bytes when combined with the 8
214 bytes of ICMP header data.
215 -S sndbuf
217 Set socket sndbuf. If not specified, it is selected to buffer not
218 more than one packet.
219 -t ttl
221 ping only. Set the IP Time to Live.
222 -T timestamp option
224 Set special IP timestamp options. timestamp option may be either
225 tsonly (only timestamps), tsandaddr (timestamps and addresses) or
226 tsprespec host1 [host2 [host3 [host4]]] (timestamp prespecified
227 hops).
228 -U
230 Print full user-to-user latency (the old behaviour). Normally ping
231 prints network round trip time, which can be different f.e. due to
232 DNS failures.
233 -v
235 Verbose output. Do not suppress DUP replies when pinging multicast
236 address.
237 -V
239 Show version and exit.
240 -w deadline
242 Specify a timeout, in seconds, before ping exits regardless of how
243 many packets have been sent or received. In this case ping does not
244 stop after count packet are sent, it waits either for deadline
245 expire or until count probes are answered or for some error
246 notification from network.
247 -W timeout
249 Time to wait for a response, in seconds. The option affects only
250 timeout in absence of any responses, otherwise ping waits for two
251 RTTs. Real number allowed with dot as a decimal separator
252 (regardless locale setup). 0 means infinite timeout.
253 When using ping for fault isolation, it should first be run on the
255 local host, to verify that the local network interface is up and
256 running. Then, hosts and gateways further and further away should be
257 “pinged”. Round-trip times and packet loss statistics are computed. If
258 duplicate packets are received, they are not included in the packet
259 loss calculation, although the round trip time of these packets is used
260 in calculating the minimum/average/maximum/mdev round-trip time
261 numbers.
262 Population standard deviation (mdev), essentially an average of how far
264 each ping RTT is from the mean RTT. The higher mdev is, the more
265 variable the RTT is (over time). With a high RTT variability, you will
266 have speed issues with bulk transfers (they will take longer than is
267 strictly speaking necessary, as the variability will eventually cause
268 the sender to wait for ACKs) and you will have middling to poor VoIP
269 quality.
270 When the specified number of packets have been sent (and received) or
272 if the program is terminated with a SIGINT, a brief summary is
273 displayed. Shorter current statistics can be obtained without
274 termination of process with signal SIGQUIT.
275 If ping does not receive any reply packets at all it will exit with
277 code 1. If a packet count and deadline are both specified, and fewer
278 than count packets are received by the time the deadline has arrived,
279 it will also exit with code 1. On other error it exits with code 2.
280 Otherwise it exits with code 0. This makes it possible to use the exit
281 code to see if a host is alive or not.
282 This program is intended for use in network testing, measurement and
284 management. Because of the load it can impose on the network, it is
285 unwise to use ping during normal operations or from automated scripts.
286
288 An IP header without options is 20 bytes. An ICMP ECHO_REQUEST packet
289 contains an additional 8 bytes worth of ICMP header followed by an
290 arbitrary amount of data. When a packetsize is given, this indicates
291 the size of this extra piece of data (the default is 56). Thus the
292 amount of data received inside of an IP packet of type ICMP ECHO_REPLY
293 will always be 8 bytes more than the requested data space (the ICMP
294 header).
295 If the data space is at least of size of struct timeval ping uses the
297 beginning bytes of this space to include a timestamp which it uses in
298 the computation of round trip times. If the data space is shorter, no
299 round trip times are given.
300
302 ping will report duplicate and damaged packets. Duplicate packets
303 should never occur, and seem to be caused by inappropriate link-level
304 retransmissions. Duplicates may occur in many situations and are rarely
305 (if ever) a good sign, although the presence of low levels of
306 duplicates may not always be cause for alarm.
307 Damaged packets are obviously serious cause for alarm and often
309 indicate broken hardware somewhere in the ping packet's path (in the
310 network or in the hosts).
311
313 The (inter)network layer should never treat packets differently
314 depending on the data contained in the data portion. Unfortunately,
315 data-dependent problems have been known to sneak into networks and
316 remain undetected for long periods of time. In many cases the
317 particular pattern that will have problems is something that doesn't
318 have sufficient “transitions”, such as all ones or all zeros, or a
319 pattern right at the edge, such as almost all zeros. It isn't
320 necessarily enough to specify a data pattern of all zeros (for example)
321 on the command line because the pattern that is of interest is at the
322 data link level, and the relationship between what you type and what
323 the controllers transmit can be complicated.
324 This means that if you have a data-dependent problem you will probably
326 have to do a lot of testing to find it. If you are lucky, you may
327 manage to find a file that either can't be sent across your network or
328 that takes much longer to transfer than other similar length files. You
329 can then examine this file for repeated patterns that you can test
330 using the -p option of ping.
331
333 The TTL value of an IP packet represents the maximum number of IP
334 routers that the packet can go through before being thrown away. In
335 current practice you can expect each router in the Internet to
336 decrement the TTL field by exactly one.
337 The TCP/IP specification states that the TTL field for TCP packets
339 should be set to 60, but many systems use smaller values (4.3 BSD uses
340 30, 4.2 used 15).
341 The maximum possible value of this field is 255, and most Unix systems
343 set the TTL field of ICMP ECHO_REQUEST packets to 255. This is why you
344 will find you can “ping” some hosts, but not reach them with telnet(1)
345 or ftp(1).
346 In normal operation ping prints the TTL value from the packet it
348 receives. When a remote system receives a ping packet, it can do one of
349 three things with the TTL field in its response:
350 • Not change it; this is what Berkeley Unix systems did before the
352 4.3BSD Tahoe release. In this case the TTL value in the received
353 packet will be 255 minus the number of routers in the round-trip
354 path.
355 • Set it to 255; this is what current Berkeley Unix systems do. In
357 this case the TTL value in the received packet will be 255 minus
358 the number of routers in the path from the remote system to the
359 pinging host.
360 • Set it to some other value. Some machines use the same value for
362 ICMP packets that they use for TCP packets, for example either 30
363 or 60. Others may use completely wild values.
364
366 • Many Hosts and Gateways ignore the RECORD_ROUTE option.
367 • The maximum IP header length is too small for options like
369 RECORD_ROUTE to be completely useful. There's not much that can be
370 done about this, however.
371 • Flood pinging is not recommended in general, and flood pinging
373 the broadcast address should only be done under very controlled
374 conditions.
375
377 ip(8), ss(8).
378
380 The ping command appeared in 4.3BSD.
381 The version described here is its descendant specific to Linux.
383 As of version s20150815, the ping6 binary doesn't exist anymore. It has
385 been merged into ping. Creating a symlink named ping6 pointing to ping
386 will result in the same functionality as before.
387
389 ping requires CAP_NET_RAW capability to be executed 1) if the program
390 is used for non-echo queries (See -N option), or 2) if kernel does not
391 support non-raw ICMP sockets, or 3) if the user is not allowed to
392 create an ICMP echo socket. The program may be used as set-uid root.
393
395 ping is part of iputils package.
396iputils 20210722 PING(8)