Scroll to navigation

SOCKET-ACTIVATE(1) User Commands SOCKET-ACTIVATE(1)

NAME

socket-activate - open sockets and exec a socket-activated service

SYNOPSIS

socket-activate [-v] [socketinfo ...] [--] prog [args]

where socketinfo is of the form:

family:[option=value,...]:address

See EXAMPLES below.

DESCRIPTION

The socket-activate program opens the requested sockets and then waits for incoming activity on them. When this happens, it execs prog, which inherits these open file descriptors. The first requested socket will be opened on file descriptor 3, and subsequent ones follow in order.

If -v or --verbose is present, socket-activate will write details of what it is doing to stderr before it execs prog.

prog is the first argument that does not start with a -, or the first argument after a --.

SOCKET FAMILIES

family can be one of several different values.

For the following values of family, if address is a single decimal number, it is treated as a port number and socket-activate will bind to all available addresses on that port. If address contains a /, the part before the / is interpreted as a network address of the requested family, and the part after the / is treated as a port number.

tcp: TCP on IPv4 and/or IPv6
udp: UDP on IPv4 and/or IPv6

For the following values of family, address is interpreted as a filesystem path.

unix: Unix-domain streaming socket
unix-dgram: Unix-domain datagram socket

SOCKET OPTIONS

Each socketinfo can have 0 or more options appended to it, separated by the comma character (,) with their value appended after an equals sign (=).

Available options are:

label: simple text string to identify this socket to the inheriting process
backlog: For streaming sockets (tcp and unix), the maximum queue for pending connections (see listen(2) – defaults to SOMAXCONN).
mode: octal mode for Unix-domain socket (defaults to relying on umask(2))
user: a numeric User ID value for Unix-domain socket ownership
group: a numeric Group ID value for Unix-domain socket ownership

EXIT CODES

socket-activate exits 100 when called with unparseable options. It prints an error message to stderr and exits 11 if it cannot successfully bind all of the requested sockets. It prints an error message to stderr and exits with the value of errno if it is unable to exec(3) the intended process. Otherwise, its exit code is the same as that of prog.

ENVIRONMENT VARIABLES

socket-activate stores the number of opened file descriptors in the environment variable $LISTEN_FDS, sets $LISTEN_PID to getpid(2), and stores the list of colon-separated labels in $LISTEN_FDNAMES.

These environment variables follow the convention understood by libsystemd (see sd_listen_fds(3)) so it should be usable with any socket-activated daemon that expects this behavior.

EXAMPLES

Run a webserver as a non-privileged user, using privileged ports, and indicating to the webserver which ports should offer HTTPS:

socket-activate --tcp::80 --tcp:label=tls:443 -- \
 runuser -u httpd -- \
  webserver

Expose a foregrounded (supervised) pulseaudio daemon to the network:

socket-activate --tcp:backlog=5:1000 -- \
 pulseaudio --daemonize=no

Run a network daemon only on the IPv4 loopback:

socket-activate --tcp::127.0.0.1/9999 -- \
 network-daemon

Run a network daemon over a unix-domain socket instead of IP:

socket-activate --unix::/srv/network-access -- \
 network-daemon

Run a public-facing network daemon with a local socket for the local administrator to control it:

socket-activate --tcp::443 --unix:label=control,user=1000,mode=0600:/run/xyz/control -- \
 xyzd

BUGS AND LIMITATIONS

socket-activate does not do any sort of name-to-number resolution: network addresses, user ids, group ids, and port numbers must all be specified numerically. Hostnames, user names, group names, service names are not interpreted.

socket-activate produces its error messages in English and is not internationalized.

socket-activate cannot provide a label that contains either a colon (:) or a comma (,).

socket-activate cannot reliably listen on a link-local network address because it has no way to specify the particular network interface.

socket-activate only opens sockets for (and launches) a single service, so it can't be used for robust system-wide synchronization. In particular, may be a race condition during startup of concurrent services.

socket-activate doesn't clearly indicate to its caller when the sockets have been successfully bound, it simply remains foregrounded.

SEE ALSO

exec(3), listen(2), runuser(1), sd_listen_fds(3), umask(2)

AUTHOR

Daniel Kahn Gillmor <dkg@fifthhorseman.net>

Feb 2019