ping — send ICMP ECHO_REQUEST to network hosts
ping [
-aAbBdCDfhHLnOqrRUvV346
] [
-c
] [
count-e
] [
identifier-F
] [
flowlabel-i
] [
interval-I
] [
interface-l
] [
preload-m
] [
mark-M
] [
pmtudisc_option-N
] [
nodeinfo_option-w
] [
deadline-W
] [
timeout-p
] [
pattern-Q
] [
tos-s
] [
packetsize-S
] [
sndbuf-t
] [
ttl-T
] [hop...] {destination}timestamp option
ping uses the ICMP protocol's mandatory ECHO_REQUEST datagram to elicit an ICMP ECHO_RESPONSE from a host or gateway. ECHO_REQUEST datagrams (“pings”) have an IP and ICMP header, followed by a struct timeval and then an arbitrary number of “pad” bytes used to fill out the packet.
ping works with both IPv4 and IPv6. Using
only one of them explicitly can be enforced by specifying
-4 or
-6.
ping can also send IPv6 Node Information Queries (RFC4620). Intermediate hops may not be allowed, because IPv6 source routing was deprecated (RFC5095).
-3
RTT precision (do not round up the result time).
-4
Use IPv4 only.
-6
Use IPv6 only.
-a
Audible ping.
-A
Adaptive ping. Interpacket interval adapts to
round-trip time, so that effectively not more than one
(or more, if preload is set) unanswered probe is present
in the network. The default interval is 2 ms,
for more info see option -i.
On networks with low RTT this mode is
essentially equivalent to flood mode.
-b
Allow pinging a broadcast address.
-B
Do not allow ping to change source address of probes. The address is bound to one selected when ping starts.
-c
count
Stop after sending count ECHO_REQUEST packets. With deadline option, ping waits for count ECHO_REPLY packets, until the timeout expires.
-C
Call connect() syscall on socket creation.
-d
Set the SO_DEBUG option on the socket being used. Essentially, this socket option is not used by Linux kernel.
-D
Print timestamp (unix time + microseconds as in gettimeofday) before each line.
-e
identifier
Set the identification field of ECHO_REQUEST.
Value 0 implies using raw socket
(not supported on ICMP datagram socket).
The value of the field may be printed with -v option.
-f
Flood ping. For every ECHO_REQUEST sent a period “.” is printed, while for every ECHO_REPLY received a backspace is printed. This provides a rapid display of how many packets are being dropped. If interval is not given, it sets interval to zero and outputs packets as fast as they come back or one hundred times per second, whichever is more. Only the super-user may use this option with zero interval.
-F
flow label
IPv6 only. Allocate and set 20 bit flow label (in hex) on echo request packets. If value is zero, kernel allocates random flow label.
-h
Show help.
-H
Force DNS name resolution for the output. Useful for numeric
destination, or -f option, which by default do not
perform it. It can also help to workaround DNS resolution problems.
Override previously defined -n option.
See also IPUTILS_PING_PTR_LOOKUP
environment variable.
-i
interval
Wait interval seconds between sending each packet. Real number allowed with dot as a decimal separator (regardless locale setup). The default is to wait for one second between each packet normally, or not to wait in flood mode. Only super-user may set interval to values less than 2 ms. Broadcast and multicast ping have even higher limitation for regular user: minimum is 1 sec.
-I
interface
interface is either an
address, an interface name or a VRF name. If
interface is an address, it
sets source address to specified interface address. If
interface is an interface
name, it sets source interface to specified interface.
If interface is a VRF
name, each packet is routed using the corresponding
routing table; in this case, the -I option
can be repeated to specify a source address.
NOTE: For IPv6, when doing ping to a link-local scope
address, link specification (by the '%'-notation in
destination, or by this
option) can be used but it is no longer required.
-l
preload
If preload is specified, ping sends that many packets not waiting for reply. Only the super-user may select preload more than 3.
-L
Suppress loopback of multicast packets. This flag only applies if the ping destination is a multicast address.
-m
mark
use mark to tag the packets going out. This is useful for variety of reasons within the kernel such as using policy routing to select specific outbound processing. CAP_NET_ADMIN or CAP_NET_RAW (since Linux 5.17) capability is required, see socket(7).
-M
pmtudisc_opt
Select Path MTU Discovery strategy. pmtudisc_option may be either do (set DF flag but subject to PMTU checks by kernel, packets too large will be rejected), want (do PMTU discovery, fragment locally when packet size is large), probe (set DF flag and bypass PMTU checks, useful for probing), or dont (do not set DF flag).
-N
nodeinfo_option
IPv6 only. Send IPv6 Node Information Queries (RFC4620), instead of Echo Request. CAP_NET_RAW capability is required.
Show help for NI support.
Queries for Node Names.
Queries for IPv6 Addresses. There are several IPv6 specific flags.
Request IPv6 global-scope addresses.
Request IPv6 site-local addresses.
Request IPv6 link-local addresses.
Request IPv6 addresses on other interfaces.
Queries for IPv4 Addresses. There is one IPv4 specific flag.
Request IPv4 addresses on other interfaces.
IPv6 subject address.
IPv4 subject address.
Subject name. If it contains more than one dot, fully-qualified domain name is assumed.
Subject name. Fully-qualified domain name is always assumed.
-n
Numeric output only. No attempt will be made to lookup
symbolic names for host addresses (no reverse DNS resolution).
This is the default for numeric destination or -f
option. Override previously defined -H option.
See also IPUTILS_PING_PTR_LOOKUP
environment variable.
-O
Report outstanding ICMP ECHO reply before sending
next packet. This is useful together with the timestamp
-D to log output to a diagnostic file and
search for missing answers.
-p
pattern
You may specify up to 16 “pad” bytes to fill out
the packet you send. This is useful for diagnosing
data-dependent problems in a network. For example,
-p ff will cause the sent packet to be
filled with all ones.
-q
Quiet output. Nothing is displayed except the summary lines at startup time and when finished.
-Q
tos
Set Quality of Service -related bits in ICMP datagrams. tos can be decimal (ping only) or hex number.
In RFC2474, these fields are interpreted as 8-bit Differentiated Services (DS), consisting of: bits 0-1 (2 lowest bits) of separate data, and bits 2-7 (highest 6 bits) of Differentiated Services Codepoint (DSCP). In RFC2481 and RFC3168, bits 0-1 are used for ECN.
Historically (RFC1349, obsoleted by RFC2474), these were interpreted as: bit 0 (lowest bit) for reserved (currently being redefined as congestion control), 1-4 for Type of Service and bits 5-7 (highest bits) for Precedence.
-r
Bypass the normal routing tables and send directly
to a host on an attached interface. If the host is not on
a directly-attached network, an error is returned. This
option can be used to ping a local host through an
interface that has no route through it provided the
option
-I is also used.
-R
ping only. Record route. Includes the RECORD_ROUTE option in the ECHO_REQUEST packet and displays the route buffer on returned packets. Note that the IP header is only large enough for nine such routes. Many hosts ignore or discard this option.
-s
packetsize
Specifies the number of data bytes to be sent. The
default is 56, which translates into 64 ICMP data bytes
when combined with the 8 bytes of ICMP header data.
The maximum allowed value is 65507 for IPv4
(65467 when -R or -T
or Intermediate hops)
or 65527 for IPv6, but most systems
limit this to a smaller, system-dependent number.
-S
sndbuf
Set socket sndbuf. If not specified, it is selected to buffer not more than one packet.
-t
ttl
ping only. Set the IP Time to Live.
-T
timestamp option
Set special IP timestamp options. timestamp option may be either tsonly (only timestamps), tsandaddr (timestamps and addresses) or tsprespec host1 [host2 [host3 [host4]]] (timestamp prespecified hops).
-U
Print full user-to-user latency (the old behaviour). Normally ping prints network round trip time, which can be different f.e. due to DNS failures.
-v
Verbose output. Do not suppress DUP replies when pinging multicast address.
-V
Show version and exit.
-w
deadline
Specify a timeout, in seconds, before ping exits regardless of how many packets have been sent or received. In this case ping does not stop after count packet are sent, it waits either for deadline expire or until count probes are answered or for some error notification from network.
-W
timeout
Time to wait for a response, in seconds. The option affects only timeout in absence of any responses, otherwise ping waits for two RTTs. Real number allowed with dot as a decimal separator (regardless locale setup). 0 means infinite timeout.
When using ping for fault isolation, it should first be run on the local host, to verify that the local network interface is up and running. Then, hosts and gateways further and further away should be “pinged”. Round-trip times and packet loss statistics are computed. If duplicate packets are received, they are not included in the packet loss calculation, although the round trip time of these packets is used in calculating the minimum/average/maximum/mdev round-trip time numbers.
Population standard deviation (mdev), essentially an average of how far each ping RTT is from the mean RTT. The higher mdev is, the more variable the RTT is (over time). With a high RTT variability, you will have speed issues with bulk transfers (they will take longer than is strictly speaking necessary, as the variability will eventually cause the sender to wait for ACKs) and you will have middling to poor VoIP quality.
When the specified number of packets have been sent (and received) or if the program is terminated with a SIGINT, a brief summary is displayed. Shorter current statistics can be obtained without termination of process with signal SIGQUIT.
This program is intended for use in network testing, measurement and management. Because of the load it can impose on the network, it is unwise to use ping during normal operations or from automated scripts.
-H or -n option.
If ping does not receive any reply packets at all it will exit with code 1. If a packet count and deadline are both specified, and fewer than count packets are received by the time the deadline has arrived, it will also exit with code 1. On other error it exits with code 2. Otherwise it exits with code 0. This makes it possible to use the exit code to see if a host is alive or not.
For IPv6, when the destination address has link-local scope and ping is using ICMP datagram sockets, the output interface must be specified. When ping is using raw sockets, it is not strictly necessary to specify the output interface but it should be done to avoid ambiguity when there are multiple possible output interfaces.
There are two ways to specify the output interface:
The destination address is postfixed with % and the output interface name or ifindex, for example:
ping fe80::5054:ff:fe70:67bc%eth0
ping fe80::5054:ff:fe70:67bc%2
When using ICMP datagram sockets, this method is supported since the following kernel versions: 5.17, 5.15.19, 5.10.96, 5.4.176, 4.19.228, 4.14.265. Also it is not supported on musl libc.
An IP header without options is 20 bytes. An ICMP ECHO_REQUEST packet contains an additional 8 bytes worth of ICMP header followed by an arbitrary amount of data. When a packetsize is given, this indicates the size of this extra piece of data (the default is 56). Thus the amount of data received inside of an IP packet of type ICMP ECHO_REPLY will always be 8 bytes more than the requested data space (the ICMP header).
If the data space is at least of size of struct timeval ping uses the beginning bytes of this space to include a timestamp which it uses in the computation of round trip times. If the data space is shorter, no round trip times are given.
ping will report duplicate and damaged packets. Duplicate packets should never occur, and seem to be caused by inappropriate link-level retransmissions. Duplicates may occur in many situations and are rarely (if ever) a good sign, although the presence of low levels of duplicates may not always be cause for alarm.
Damaged packets are obviously serious cause for alarm and often indicate broken hardware somewhere in the ping packet's path (in the network or in the hosts).
Unlike TCP and UDP, which use port to uniquely identify the recipient to deliver data, ICMP uses identifier field (ID) for identification. Therefore, if on the same machine, at the same time, two ping processes use the same ID, echo reply can be delivered to a wrong recipient. This is a known problem due to the limited size of the 16-bit ID field. That is a historical limitation of the protocol that cannot be fixed at the moment unless we encode an ID into the ping packet payload. ping prints DIFFERENT ADDRESS error and packet loss is negative.
ping uses PID to get unique number. The default value of /proc/sys/kernel/pid_max is 32768. On the systems that use ping heavily and with pid_max greater than 65535 collisions are bound to happen.
The (inter)network layer should never treat packets differently depending on the data contained in the data portion. Unfortunately, data-dependent problems have been known to sneak into networks and remain undetected for long periods of time. In many cases the particular pattern that will have problems is something that doesn't have sufficient “transitions”, such as all ones or all zeros, or a pattern right at the edge, such as almost all zeros. It isn't necessarily enough to specify a data pattern of all zeros (for example) on the command line because the pattern that is of interest is at the data link level, and the relationship between what you type and what the controllers transmit can be complicated.
This means that if you have a data-dependent problem you
will probably have to do a lot of testing to find it. If you
are lucky, you may manage to find a file that either can't be
sent across your network or that takes much longer to transfer
than other similar length files. You can then examine this file
for repeated patterns that you can test using the
-p option of
ping.
The TTL value of an IP packet represents the maximum number of IP routers that the packet can go through before being thrown away. In current practice you can expect each router in the Internet to decrement the TTL field by exactly one.
The TTL field for TCP packets may take various values. The maximum possible value of this field is 255, a recommended initial value is 64. For more information, see the TCP/Lower-Level Interface section of RFC9293.
In normal operation ping prints the TTL value from the packet it receives. When a remote system receives a ping packet, it can do one of three things with the TTL field in its response:
• Not change it; this is what Berkeley Unix systems did before the 4.3BSD Tahoe release. In this case the TTL value in the received packet will be 255 minus the number of routers in the round-trip path.
• Set it to 255; this is what current Berkeley Unix systems do. In this case the TTL value in the received packet will be 255 minus the number of routers in the path from the remote system to the pinging host.
• Set it to some other value. Some machines use the same value for ICMP packets that they use for TCP packets, for example either 30 or 60. Others may use completely wild values.
• Many Hosts and Gateways ignore the RECORD_ROUTE option.
• The maximum IP header length is too small for options like RECORD_ROUTE to be completely useful. There's not much that can be done about this, however.
• Flood pinging is not recommended in general, and flood pinging the broadcast address should only be done under very controlled conditions.
The ping command appeared in 4.3BSD.
The version described here is its descendant specific to Linux.
As of version s20150815, the ping6 binary doesn't exist anymore. It has been merged into ping. Creating a symlink named ping6 pointing to ping will result in the same functionality as before.
ping requires CAP_NET_RAW capability to be
executed 1) if the program is used for non-echo queries (see
-N option) or when the identification field set to 0
for ECHO_REQUEST (see -e), or
2) if kernel does not support ICMP datagram sockets, or
3) if the user is not allowed to create an ICMP echo socket.
The program may be used as set-uid root.