The s6 command
s6-frontend
Software
skarnet.org

The s6 process command

s6 process groups actions that target services that are managed by an instance of s6-svscan running on the configured scandir. All commands (other than help) require the services in servicenames... to be a valid reference to a service directory with an instance of s6-supervise supervising it.

The point of s6 process is to make it easier to send commands to supervised services, by not requiring the user to know where the scandir is or the details of . Instead of s6-svc -r /run/service/foobar, the user can type s6 process restart foobar for the same effect.

Interface

     s6 process subcommand [ subcommand_options... ] [ args... ]

Subcommands

help

s6 process help prints a short help message summarizing the options and usage of the s6 process command. It is not as detailed as this page.

kill

Interface

     s6 process kill [ -W | -w ] [ -s sig ] [ -t timeout ] servicenames...

• s6 process kill servicenames... sends a signal to every supervised daemon listed by name in servicenames.
• Note that if the daemon dies as a consequence of receiving the signal, it will normally be restarted by its supervisor. To kill a daemon and keep it from restarting you need to use s6 process stop.

Options

-W, --no-wait

Exit immediately, do not wait until the daemon has died and the service has cleaned up. This is the default.

-w, --wait

Only exit when the daemon has died and the finish script of the service, if any, has run.

-s sig, --signal=sig

Send signal sig. The default is SIGTERM.

-t timeout, --timeout=timeout

Only wait for at most timeout milliseconds; if the daemon still hasn't died after that time, exit. The default is 0, meaning infinite: the command can wait forever. This option is only effective when paired with the --wait option.

status

Interface

     s6 process status [ -L | -l ] servicenames...

• s6 process status servicenames... prints a line of status for every supervised daemon listed by name in servicenames.

Options

-L, --without-logs

Only print the status for /run/service/foobar even if /run/service/foobar/log exists. This is the default.

-l, --with-logs

If a /run/service/foobar/log service directory exists, also print its status along with the one for /run/service/foobar. Note that this option has no effect on loggers managed by s6-rc, e.g. services defined in the store and managed by s6 live, because these loggers are treated at separate, independent services. The -l option is only useful with old-style logged s6 services that are not managed by s6-rc.

start

Interface

     s6 process start [ -P | -p ] [ -W | -w ] [ -t timeout ] servicenames...

• s6 process start servicenames... brings up every supervised daemon listed by name in servicenames.

Options

-P, --no-permanent

Only change the current state of the service, not its default state. If there is a down file in /run/service/foobar, then s6 process start foobar will bring the service up, but it will not be restarted if foobard dies. This is the default.

-p, --permanent

Change the current and default states of the service. If there is a down file in /run/service/foobar, then s6 process start foobar will delete it, and the service will be automatically be restarted if foobard dies.

-W, --no-wait

Exit immediately, do not wait until the service is up. This is the default.

-w, --wait

Only exit when the service is up and (if applicable) ready.

-t timeout, --timeout=timeout

Only wait for at most timeout milliseconds; if the daemon still isn't up and ready after that time, exit. The default is 0, meaning infinite: the command can wait forever. This option is only effective when paired with the --wait option.

stop

Interface

     s6 process stop [ -P | -p ] [ -W | -w ] [ -t timeout ] servicenames...

• s6 process stop servicenames... brings down every supervised daemon listed by name in servicenames.

Options

-P, --no-permanent

Only change the current state of the service, not its default state. If there is no down file in /run/service/foobar, then s6 process stop foobar will stop the service, but it will still be started whenever its supervisor is started. This is the default.

-p, --permanent

Change the current and default states of the service. If there is no down file in /run/service/foobar, then s6 process stop foobar will create one, and the service will not be automatically started whenever its supervisor is started.

-W, --no-wait

Exit immediately, do not wait until the service is down. This is the default.

-w, --wait

Only exit when the service is down and any cleanups are done.

-t timeout, --timeout=timeout

Only wait for at most timeout milliseconds; if the finish script still isn't done after that time, exit. The default is 0, meaning infinite: the command can wait forever. This option is only effective when paired with the --wait option.

restart

Interface

     s6 process restart [ -W | -w ] [ -t timeout ] servicenames...

• s6 process restart servicenames... restarts every supervised daemon listed by name in servicenames. This is achieved by issuing a stop followed by a start command on the targeted service(s).

Options

-W, --no-wait

Exit immediately, do not wait until the service has restarted This is the default.

-w, --wait

Only exit when the new instance of the service is up and (if applicable) ready.

-t timeout, --timeout=timeout

Only wait for at most timeout milliseconds; if the service still hasn't restarted after that time, exit. The default is 0, meaning infinite: the command can wait forever. This option is only effective when paired with the --wait option.