View Source gen_tcp (kernel v10.1.1)
Interface to TCP/IP sockets.
This module provides functions for communicating over TCP/IP protocol sockets.
The following code fragment is a simple example of a client connecting to a server at port 5678, transferring a binary, and closing the connection:
client() ->
SomeHostInNet = "localhost", % to make it runnable on one machine
{ok, Sock} = gen_tcp:connect(SomeHostInNet, 5678,
[binary, {packet, 0}]),
ok = gen_tcp:send(Sock, "Some Data"),
ok = gen_tcp:close(Sock).
At the other end, a server is listening on port 5678, accepts the connection, and receives the binary:
server() ->
{ok, LSock} = gen_tcp:listen(5678, [binary, {packet, 0},
{active, false}]),
{ok, Sock} = gen_tcp:accept(LSock),
{ok, Bin} = do_recv(Sock, []),
ok = gen_tcp:close(Sock),
ok = gen_tcp:close(LSock),
Bin.
do_recv(Sock, Bs) ->
case gen_tcp:recv(Sock, 0) of
{ok, B} ->
do_recv(Sock, [Bs, B]);
{error, closed} ->
{ok, list_to_binary(Bs)}
end.
For more examples, see section Examples.
Note
Functions that create sockets can take an optional option;
{inet_backend, Backend}
that, if specified, has to be the first option. This selects the implementation backend towards the platform's socket API.This is a temporary option that will be ignored in a future release.
The default is
Backend = inet
that selects the traditionalinet_drv.c
driver. The other choice isBackend = socket
that selects the newsocket
module and its NIF implementation.The system default can be changed when the node is started with the application
kernel
's configuration variableinet_backend
.For
gen_tcp
withinet_backend = socket
we have tried to be as "compatible" as possible which has sometimes been impossible. Here is a list of cases when the behaviour of inet-backendinet
(default) andsocket
are different:
If a user calling
gen_tcp:send/2
withinet_backend = inet
, tries to send more data than there is room for in the OS buffers, the "rest data" is buffered by the inet driver (and later sent in the background). The effect for the user is that the call is non-blocking.This is not the effect when
inet_backend = socket
, since there is no buffering. Instead the user hangs either until all data has been sent or thesend_timeout
timeout has been reached.
shutdown/2
may hide errorsThe call does not involve the receive process state, and is done right on the underlying socket. On for example Linux, it is a known misbehaviour that it skips some checks so doing shutdown on a listen socket returns
ok
while the logical result should have been{error, enotconn}
. Theinet_drv.c
driver did an extra check and simulated the correct error, but withBackend = socket
it would introduce overhead to involve the receive process.The option nodelay is a TCP specific option that is not compatible with
domain = local
.When using
inet_backend = socket
, trying to create a socket (via listen or connect) withdomain = local
(for example with option {ifaddr, {local,"/tmp/test"}}) will fail with{error, enotsup}
.This does not actually work for
inet_backend = inet
either, but in that case the error is simply ignored, which is a bad idea. We have chosen to not ignore this error forinet_backend = socket
.Calling gen_tcp:shutdown(Socket, write | read_write) on a socket created with
inet_backend = socket
will take immediate effect, unlike for a socket created withinet_backend = inet
.See async shutdown write for more info.
Windows require sockets (domain =
inet | inet6
) to be bound.Currently all sockets created on Windows with
inet_backend = socket
will be bound. If the user does not provide an address, gen_tcp will try to 'figure out' an address itself.
Examples
The following example illustrates use of option {active,once}
and multiple
accepts by implementing a server as a number of worker processes doing accept on
a single listening socket. Function start/2
takes the number of worker
processes and the port number on which to listen for incoming connections. If
LPort
is specified as 0
, an ephemeral port number is used, which is why the
start function returns the actual port number allocated:
start(Num,LPort) ->
case gen_tcp:listen(LPort,[{active, false},{packet,2}]) of
{ok, ListenSock} ->
start_servers(Num,ListenSock),
{ok, Port} = inet:port(ListenSock),
Port;
{error,Reason} ->
{error,Reason}
end.
start_servers(0,_) ->
ok;
start_servers(Num,LS) ->
spawn(?MODULE,server,[LS]),
start_servers(Num-1,LS).
server(LS) ->
case gen_tcp:accept(LS) of
{ok,S} ->
loop(S),
server(LS);
Other ->
io:format("accept returned ~w - goodbye!~n",[Other]),
ok
end.
loop(S) ->
inet:setopts(S,[{active,once}]),
receive
{tcp,S,Data} ->
Answer = process(Data), % Not implemented in this example
gen_tcp:send(S,Answer),
loop(S);
{tcp_closed,S} ->
io:format("Socket ~w closed [~w]~n",[S,self()]),
ok
end.
Example of a simple client:
client(PortNo,Message) ->
{ok,Sock} = gen_tcp:connect("localhost",PortNo,[{active,false},
{packet,2}]),
gen_tcp:send(Sock,Message),
A = gen_tcp:recv(Sock,0),
gen_tcp:close(Sock),
A.
The send
call does not accept a time-out option because time-outs on send is
handled through socket option send_timeout
. The behavior of a send operation
with no receiver is mainly defined by the underlying TCP stack and the network
infrastructure. To write code that handles a hanging receiver that can
eventually cause the sender to hang on a send
do like the following.
Consider a process that receives data from a client process to be forwarded to a
server on the network. The process is connected to the server through TCP/IP and
does not get any acknowledge for each message it sends, but has to rely on the
send time-out option to detect that the other end is unresponsive. Option
send_timeout
can be used when connecting:
...
{ok,Sock} = gen_tcp:connect(HostAddress, Port,
[{active,false},
{send_timeout, 5000},
{packet,2}]),
loop(Sock), % See below
...
In the loop where requests are handled, send time-outs can now be detected:
loop(Sock) ->
receive
{Client, send_data, Binary} ->
case gen_tcp:send(Sock,[Binary]) of
{error, timeout} ->
io:format("Send timeout, closing!~n",
[]),
handle_send_timeout(), % Not implemented here
Client ! {self(),{error_sending, timeout}},
%% Usually, it's a good idea to give up in case of a
%% send timeout, as you never know how much actually
%% reached the server, maybe only a packet header?!
gen_tcp:close(Sock);
{error, OtherSendError} ->
io:format("Some other error on socket (~p), closing",
[OtherSendError]),
Client ! {self(),{error_sending, OtherSendError}},
gen_tcp:close(Sock);
ok ->
Client ! {self(), data_sent},
loop(Sock)
end
end.
Usually it suffices to detect time-outs on receive, as most protocols include
some sort of acknowledgment from the server, but if the protocol is strictly one
way, option send_timeout
comes in handy.
Summary
Functions
Equivalent to accept(ListenSocket, infinity)
.
Accept an incoming connection request on a listen socket.
Close a TCP socket.
Equivalent to connect(SockAddr, Opts, infinity)
.
Create a socket connected to the specified address.
Create a socket connected to the specified address.
Change the controlling process (owner) of a socket.
Create a listen socket.
Equivalent to recv(Socket, Length, infinity)
.
Receive a packet, from a socket in passive mode.
Send a packet on a socket.
Close the socket in one or both directions.
Types
-type connect_option() :: {fd, Fd :: non_neg_integer()} | inet:address_family() | {ifaddr, socket:sockaddr_in() | socket:sockaddr_in6() | inet:socket_address()} | {ip, inet:socket_address()} | {port, inet:port_number()} | {tcp_module, module()} | {netns, file:filename_all()} | {bind_to_device, binary()} | option().
-type listen_option() :: {fd, Fd :: non_neg_integer()} | inet:address_family() | {ifaddr, socket:sockaddr_in() | socket:sockaddr_in6() | inet:socket_address()} | {ip, inet:socket_address()} | {port, inet:port_number()} | {backlog, B :: non_neg_integer()} | {tcp_module, module()} | {netns, file:filename_all()} | {bind_to_device, binary()} | option().
-type option() :: {active, true | false | once | -32768..32767} | {buffer, non_neg_integer()} | {debug, boolean()} | {delay_send, boolean()} | {deliver, port | term} | {dontroute, boolean()} | {exit_on_close, boolean()} | {exclusiveaddruse, boolean()} | {header, non_neg_integer()} | {high_msgq_watermark, pos_integer()} | {high_watermark, non_neg_integer()} | {keepalive, boolean()} | {linger, {boolean(), non_neg_integer()}} | {low_msgq_watermark, pos_integer()} | {low_watermark, non_neg_integer()} | {mode, list | binary} | list | binary | {nodelay, boolean()} | {packet, 0 | 1 | 2 | 4 | raw | sunrm | asn1 | cdr | fcgi | line | tpkt | http | httph | http_bin | httph_bin} | {packet_size, non_neg_integer()} | {priority, non_neg_integer()} | {raw, Protocol :: non_neg_integer(), OptionNum :: non_neg_integer(), ValueBin :: binary()} | {recbuf, non_neg_integer()} | {reuseaddr, boolean()} | {reuseport, boolean()} | {reuseport_lb, boolean()} | {send_timeout, non_neg_integer() | infinity} | {send_timeout_close, boolean()} | {show_econnreset, boolean()} | {sndbuf, non_neg_integer()} | {tos, non_neg_integer()} | {tclass, non_neg_integer()} | {ttl, non_neg_integer()} | {recvtos, boolean()} | {recvtclass, boolean()} | {recvttl, boolean()} | {ipv6_v6only, boolean()}.
-type option_name() :: active | buffer | debug | delay_send | deliver | dontroute | exit_on_close | exclusiveaddruse | header | high_msgq_watermark | high_watermark | keepalive | linger | low_msgq_watermark | low_watermark | mode | nodelay | packet | packet_size | priority | {raw, Protocol :: non_neg_integer(), OptionNum :: non_neg_integer(), ValueSpec :: (ValueSize :: non_neg_integer()) | (ValueBin :: binary())} | recbuf | reuseaddr | reuseport | reuseport_lb | send_timeout | send_timeout_close | show_econnreset | sndbuf | tos | tclass | ttl | recvtos | recvtclass | recvttl | pktoptions | ipv6_v6only.
-type pktoptions_value() :: {pktoptions, inet:ancillary_data()}.
Value from socket option pktoptions
.
If the platform implements the IPv4 option IP_PKTOPTIONS
,
or the IPv6 option IPV6_PKTOPTIONS
or IPV6_2292PKTOPTIONS
for the socket;
this value is returned from inet:getopts/2
when called with the option name
pktoptions
.
Note
This option appears to be VERY Linux specific, and its existence in future Linux kernel versions is also worrying since the option is part of RFC 2292 which is since long (2003) obsoleted by RFC 3542 that explicitly removes this possibility to get packet information from a stream socket. For comparison: it has existed in FreeBSD but is now removed, at least since FreeBSD 10.
-type socket() :: inet:socket().
As returned by accept/1,2
and connect/3,4
.
Functions
-spec accept(ListenSocket) -> {ok, Socket} | {error, Reason} when ListenSocket :: socket(), Socket :: socket(), Reason :: closed | system_limit | inet:posix().
Equivalent to accept(ListenSocket, infinity)
.
-spec accept(ListenSocket, Timeout) -> {ok, Socket} | {error, Reason} when ListenSocket :: socket(), Timeout :: timeout(), Socket :: socket(), Reason :: closed | timeout | system_limit | inet:posix().
Accept an incoming connection request on a listen socket.
Socket
must be a socket returned from listen/2
. Timeout
specifies
a time-out value in milliseconds. Defaults to infinity
.
Returns:
{ok, Socket}
if a connection is established{error, closed}
ifListenSocket
is closed{error, timeout}
if no connection is established withinTimeout
{error, system_limit}
if all available ports in the Erlang emulator are in use- A POSIX error value if something else goes wrong, see
inet
about possible values
To send packets (outbound) on the returned Socket
, use send/2
.
Packets sent from the peer (inbound) are delivered as messages
to the socket owner; the process that created the socket.
Unless {active, false}
is specified in the option list when creating
the listening socket.
See connect/4
about active mode socket messages and passive mode.
Note
The
accept
call doesn't have to be issued from the socket owner process. Using version 5.5.3 and higher of the emulator, multiple simultaneous accept calls can be issued from different processes, which allows for a pool of acceptor processes handling incoming connections.
-spec close(Socket) -> ok when Socket :: socket().
Close a TCP socket.
Note that in most implementations of TCP, doing a close
does not guarantee
that the data sent is delivered to the recipient. It is guaranteed that
the recepient will see all sent data before getting the close, but the
sender gets no indication of that.
If the sender needs to know that the recepient has received all data there are two common ways to achieve this:
- Use
gen_tcp:shutdown(Sock, write)
to signal that no more data is to be sent and wait for the other side to acknowledge seeing its read side being closed, by closing its write side, which shows as a socket close on this side. - Implement an acknowledgement in the protocol on top of TCP
that both connection ends adhere to, indicating that all data
has been seen. The socket option
{packet, N}
may be useful.
-spec connect(SockAddr, Opts) -> {ok, Socket} | {error, Reason} when SockAddr :: socket:sockaddr_in() | socket:sockaddr_in6(), Opts :: [inet:inet_backend() | connect_option()], Socket :: socket(), Reason :: inet:posix().
Equivalent to connect(SockAddr, Opts, infinity)
.
-spec connect(Address, Port, Opts) -> {ok, Socket} | {error, Reason} when Address :: inet:socket_address() | inet:hostname(), Port :: inet:port_number(), Opts :: [inet:inet_backend() | connect_option()], Socket :: socket(), Reason :: inet:posix(); (SockAddr, Opts, Timeout) -> {ok, Socket} | {error, Reason} when SockAddr :: socket:sockaddr_in() | socket:sockaddr_in6(), Opts :: [inet:inet_backend() | connect_option()], Timeout :: timeout(), Socket :: socket(), Reason :: timeout | inet:posix().
Create a socket connected to the specified address.
With arguments Address
and Port
Equivalent to connect(Address, Port, Opts, infinity)
.
With argument SockAddr
(since OTP 24.3)
Connects to a remote listen socket specified by SockAddr
where socket:sockaddr_in6/0
for example allows specifying
the scope_id
for link local IPv6 addresses.
IPv4 addresses on the same
map/0
format is also allowed.
Equivalent to connect/4
, besides the format of the destination address.
-spec connect(Address, Port, Opts, Timeout) -> {ok, Socket} | {error, Reason} when Address :: inet:socket_address() | inet:hostname(), Port :: inet:port_number(), Opts :: [inet:inet_backend() | connect_option()], Timeout :: timeout(), Socket :: socket(), Reason :: timeout | inet:posix().
Create a socket connected to the specified address.
Creates a socket and connects it to a server on TCP port Port
on the host with IP address Address
, that may also be a hostname.
Opts
(connect options)
{ip, Address}
- If the local host has many IP addresses, this option specifies which one to use.{ifaddr, Address}
- Same as{ip, Address}
.However, if
Address
instead is asocket:sockaddr_in/0
orsocket:sockaddr_in6/0
this takes precedence over any value previously set with theip
andport
options. If these options (ip
or/andport
) however comes after this option, they may be used to update the corresponding fields of this option (forip
, theaddr
field, and forport
, theport
field).{fd, integer() >= 0}
- If a socket has somehow been connected without usinggen_tcp
, use this option to pass the file descriptor for it. If{ip, Address}
and/or{port, port_number()}
is combined with this option, thefd
is bound to the specified interface and port before connecting. If these options are not specified, it is assumed that thefd
is already bound appropriately.inet
- Sets up the socket for IPv4.inet6
- Sets up the socket for IPv6.local
- Sets up a Unix Domain Socket. Seeinet:local_address/0
{port, Port}
- Specifies which local port number to use.{tcp_module, module()}
- Overrides which callback module is used. Defaults toinet_tcp
for IPv4 andinet6_tcp
for IPv6.option/0
- Seeinet:setopts/2
.
Socket Data
Packets can be sent to the peer (outbound) with
send(Socket, Packet)
. Packets sent from the peer
(inbound) are delivered as messages to the socket owner;
the process that created the socket, unless {active, false}
is specified in the Options
list.
Active mode socket messages
{tcp, Socket, Data}
- Inbound data from the socket.{tcp_passive, Socket}
- The socket was in{active, N}
mode (seeinet:setopts/2
for details) and its message counter reached0
, indicating that the socket has transitioned to passive ({active, false}
) mode.{tcp_closed, Socket}
- The socket was closed.{tcp_error, Socket, Reason}
A socket error occurred.
Passive mode
If {active, false}
is specified in the option list for the socket,
packets and errors are retrieved by calling recv/2,3
(send/2
may also return errors).
Timeout
The optional Timeout
parameter specifies a connect time-out in milliseconds.
Defaults to infinity
.
Note
Keep in mind that if the underlying OS
connect()
call returns a timeout,gen_tcp:connect
will also return a timeout (i.e.{error, etimedout}
), even if a largerTimeout
was specified (for exampleinfinity
).
Note
The default values for options specified to
connect
can be affected by the Kernel configuration parameterinet_default_connect_options
. For details, seeinet
.
-spec controlling_process(Socket, Pid) -> ok | {error, Reason} when Socket :: socket(), Pid :: pid(), Reason :: closed | not_owner | badarg | inet:posix().
Change the controlling process (owner) of a socket.
Assigns a new controlling process Pid
to Socket
. The controlling process
is the process that the socket sends messages to. If this function
is called from any other process than the current controlling process,
{error, not_owner}
is returned.
If the process identified by Pid
is not an existing local pid/0
,
{error, badarg}
is returned. {error, badarg}
may also be returned
in some cases when Socket
is closed during the execution of this function.
If the socket is in active mode, this function will transfer any messages from the socket in the mailbox of the caller to the new controlling process.
If any other process is interacting with the socket during the transfer, it may not work correctly and messages may remain in the caller's mailbox. For instance, changing the sockets active mode during the transfer could cause this.
-spec listen(Port, Options) -> {ok, ListenSocket} | {error, Reason} when Port :: inet:port_number(), Options :: [inet:inet_backend() | listen_option()], ListenSocket :: socket(), Reason :: system_limit | inet:posix().
Create a listen socket.
Creates a socket and sets it to listen on port Port
on the local host.
If Port == 0
, the underlying OS assigns an available (ephemeral)
port number, use inet:port/1
to retrieve it.
The following options are available:
list
- ReceivedPacket
s are delivered as lists of bytes,[
byte/0
]
.binary
- ReceivedPacket
s are delivered asbinary/0
s.{backlog, B}
-B ::
non_neg_integer/0
. The backlog value defines the maximum length that the queue of pending connections can grow to. Defaults to5
.inet6
- Sets up the socket for IPv6.inet
- Sets up the socket for IPv4.{fd, Fd}
- If a socket has somehow been created without usinggen_tcp
, use this option to pass the file descriptor for it.{ip, Address}
- If the host has many IP addresses, this option specifies which one to listen on.{port, Port}
- Specifies which local port number to use.{ifaddr, Address}
- Same as{ip, Address}
.However, if this instead is an
socket:sockaddr_in/0
orsocket:sockaddr_in6/0
this takes precedence over any value previously set with theip
andport
options. If these options (ip
or/andport
) however comes after this option, they may be used to update their corresponding fields of this option (forip
, theaddr
field, and forport
, theport
field).{tcp_module, module()}
- Overrides which callback module is used. Defaults toinet_tcp
for IPv4 andinet6_tcp
for IPv6.option/0
- Seeinet:setopts/2
.
The returned socket ListenSocket
should be used when calling
accept/1,2
to accept an incoming connection request.
Note
The default values for options specified to
listen
can be affected by the Kernel configuration parameterinet_default_listen_options
. For details, seeinet
.
-spec recv(Socket, Length) -> {ok, Packet} | {error, Reason} when Socket :: socket(), Length :: non_neg_integer(), Packet :: string() | binary() | HttpPacket, Reason :: closed | inet:posix(), HttpPacket :: term().
Equivalent to recv(Socket, Length, infinity)
.
-spec recv(Socket, Length, Timeout) -> {ok, Packet} | {error, Reason} when Socket :: socket(), Length :: non_neg_integer(), Timeout :: timeout(), Packet :: string() | binary() | HttpPacket, Reason :: closed | timeout | inet:posix(), HttpPacket :: term().
Receive a packet, from a socket in passive mode.
A closed socket is indicated by the return value {error, closed}
.
If the socket is not in passive mode, the return value is {error, einval}
.
Argument Length
is only meaningful when the socket is in raw
mode and
denotes the number of bytes to read. If Length
is 0
, all available
bytes are returned. If Length > 0
, exactly Length
bytes are returned,
or an error; except if the socket is closed from the other side,
then the last read before the one returning {error, closed}
may return less than Length
bytes of data.
The optional Timeout
parameter specifies a time-out in milliseconds.
Defaults to infinity
.
Any process can receive data from a passive socket, even if that process is not
the controlling process of the socket. However, only one process can call this
function on a socket at any given time. Using simultaneous calls to recv
is
not recommended as the behavior depends on the socket implementation,
and could return errors such as {error, ealready}
.
-spec send(Socket, Packet) -> ok | {error, Reason} when Socket :: socket(), Packet :: iodata(), Reason :: closed | {timeout, RestData} | inet:posix(), RestData :: binary() | erlang:iovec().
Send a packet on a socket.
There is no send/2
call with a time-out option; use socket option
send_timeout
if time-outs are desired. See section
Examples.
The return value {error, {timeout, RestData}}
can only be returned when
inet_backend = socket
.
Note
Non-blocking send.
If the user tries to send more data than there is room for in the OS send buffers, the 'rest data' is stored in (inet driver) internal buffers and later sent in the background. The function immediately returns ok (not informing the caller that some date isn'nt sent yet). Any issue while sending the 'rest data' may be returned later.
When using
inet_backend = socket
, the behaviour is different. There is no buffering, instead the caller will "hang" until all of the data has been sent or the send timeout (as specified by thesend_timeout
option) expires (the function can "hang" even when using theinet
backend if the internal buffers are full).If this happens when using
packet =/= raw
, a partial package has been written. A new package therefore mustn't be written at this point, as there is no way for the peer to distinguish this from data in the current package. Instead, set package to raw, send the rest data (as raw data) and then set package to the correct package type again.
-spec shutdown(Socket, How) -> ok | {error, Reason} when Socket :: socket(), How :: read | write | read_write, Reason :: inet:posix().
Close the socket in one or both directions.
How == write
means closing the socket for writing, reading from it is still
possible.
If How == read
or there is no outgoing data buffered in the Socket
port, the
shutdown is performed immediately and any error encountered is returned in
Reason
.
If there is data buffered in the socket port, shutdown isn't performed
on the socket until that buffered data has been written to the OS
protocol stack. If any errors are encountered, the socket is closed
and {error, closed}
is returned by the next recv/2
or send/2
call.
Option {exit_on_close, false}
is useful if the peer performs a shurdown
of its write side. Then the socket stays open for writing after
receive has indicated that the socket was closed.
Note
Async shutdown write (
How :: write | read_write
).If the shutdown attempt is made while the inet driver is sending buffered data in the background, the shutdown is postponed until all buffered data has been sent. This function immediately returns
ok
, and the caller isn't informed (that the shutdown has been postponed).When using
inet_backend = socket
, the behaviour is different. A shutdown withHow :: write | read_write
will always be performed immediately.