Copyright 2010 Lennart Poettering
systemd is free software; you can redistribute it and/or modify it
- under the terms of the GNU General Public License as published by
- the Free Software Foundation; either version 2 of the License, or
+ under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2.1 of the License, or
(at your option) any later version.
systemd is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- General Public License for more details.
+ Lesser General Public License for more details.
- You should have received a copy of the GNU General Public License
+ You should have received a copy of the GNU Lesser General Public License
along with systemd; If not, see <http://www.gnu.org/licenses/>.
-->
<refnamediv>
<refname>systemd.service</refname>
- <refpurpose>systemd service configuration files</refpurpose>
+ <refpurpose>Service unit configuration</refpurpose>
</refnamediv>
<refsynopsisdiv>
<para>Additional options are listed in
<citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
which define the execution environment the commands
- are executed in.</para>
+ are executed in, and in
+ <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+ which define the way the processes of the service are
+ terminated.</para>
<para>Unless <varname>DefaultDependencies=</varname>
is set to <option>false</option>, service units will
<filename>.service</filename> suffix removed) and
dynamically creates a service unit from that
script. This is useful for compatibility with
- SysV.</para>
+ SysV. Note that this compatibility is quite
+ comprehensive but not 100%. For details about the
+ incomptibilities see the <ulink
+ url="http://www.freedesktop.org/wiki/Software/systemd/Incompatibilities">Incompatibilities
+ with SysV</ulink> document.
+ </para>
</refsect1>
<refsect1>
supervises. A number of options that may be used in
this section are shared with other unit types. These
options are documented in
- <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>. The
+ <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+ and
+ <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>. The
options specific to the <literal>[Service]</literal>
section of service units are the following:</para>
<option>forking</option>,
<option>oneshot</option>,
<option>dbus</option>,
- <option>notify</option>.</para>
+ <option>notify</option> or
+ <option>idle</option>.</para>
<para>If set to
<option>simple</option> (the default
- value) it is expected that the process
- configured with
+ value if <varname>BusName=</varname>
+ is not specified) it is expected that
+ the process configured with
<varname>ExecStart=</varname> is the
main process of the service. In this
mode, if the process offers
complete and all communication
channels set up. The child continues
to run as the main daemon
- process. This is the behaviour of
+ process. This is the behavior of
traditional UNIX daemons. If this
setting is used, it is recommended to
also use the
soon as the parent process
exits.</para>
- <para>Behaviour of
+ <para>Behavior of
<option>oneshot</option> is similar
to <option>simple</option>, however
it is expected that the process has to
is particularly useful for this type
of service.</para>
- <para>Behaviour of
+ <para>Behavior of
<option>dbus</option> is similar to
<option>simple</option>, however it is
expected that the daemon acquires a
acquired. Service units with this
option configured implicitly gain
dependencies on the
- <filename>dbus.target</filename>
- unit.</para>
+ <filename>dbus.socket</filename>
+ unit. This type is the default if
+ <varname>BusName=</varname> is
+ specified.</para>
- <para>Behaviour of
+ <para>Behavior of
<option>notify</option> is similar to
<option>simple</option>, however it is
expected that the daemon sends a
below) should be set to open access to
the notification socket provided by
systemd. If
- <varname>NotifyAccess=</varname> is not
- set, it will implicitly be set to
+ <varname>NotifyAccess=</varname> is
+ not set, it will be implicitly set to
<option>main</option>.</para>
+
+ <para>Behavior of
+ <option>idle</option> is very similar
+ to <option>simple</option>, however
+ actual execution of a the service
+ binary is delayed until all jobs are
+ dispatched. This may be used to avoid
+ interleaving of output of shell
+ services with the status output on the
+ console.</para>
</listitem>
</varlistentry>
</listitem>
</varlistentry>
+ <varlistentry>
+ <term><varname>GuessMainPID=</varname></term>
+
+ <listitem><para>Takes a boolean value
+ that specifies whether systemd should
+ try to guess the main PID of a service
+ should if it cannot be determined
+ reliably. This option is ignored
+ unless <option>Type=forking</option>
+ is set and <option>PIDFile=</option>
+ is unset because for the other types
+ or with an explicitly configured PID
+ file the main PID is always known. The
+ guessing algorithm might come to
+ incorrect conclusions if a daemon
+ consists of more than one process. If
+ the main PID cannot be determined
+ failure detection and automatic
+ restarting of a service will not work
+ reliably. Defaults to
+ <option>yes</option>.</para>
+ </listitem>
+ </varlistentry>
+
<varlistentry>
<term><varname>PIDFile=</varname></term>
daemon. Use of this option is
recommended for services where
<varname>Type=</varname> is set to
- <option>forking</option>.</para>
+ <option>forking</option>. systemd will
+ read the PID of the main process of
+ the daemon after start-up of the
+ service. systemd will not write to the
+ file configured here.</para>
</listitem>
</varlistentry>
<literal>-</literal> an exit code of
the command normally considered a
failure (i.e. non-zero exit status or
- abormal exit due to signal) is ignored
+ abnormal exit due to signal) is ignored
and considered success. If both
<literal>-</literal> and
<literal>@</literal> are used for the
- same command the former must preceed
+ same command the former must precede
the latter. Unless
<varname>Type=forking</varname> is
set, the process started via this
main process of the daemon. The
command line accepts % specifiers as
described in
- <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>. On
- top of that basic environment variable
- substitution is supported, where
- <literal>${FOO}</literal> is replaced
- by the string value of the environment
- variable of the same name. Also
- <literal>$FOO</literal> may appear as
- seperate word on the command line in
- which case the variable is replaced by
- its value split at
- whitespaces.</para></listitem>
+ <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para>
+
+ <para>On top of that basic environment
+ variable substitution is
+ supported. Use
+ <literal>${FOO}</literal> as part of a
+ word, or as word of its own on the
+ command line, in which case it will be
+ replaced by the value of the
+ environment variable including all
+ whitespace it contains, resulting in a
+ single argument. Use
+ <literal>$FOO</literal> as a separate
+ word on the command line, in which
+ case it will be replaced by the value
+ of the environment variable split up
+ at whitespace, resulting in no or more
+ arguments. Note that the first
+ argument (i.e. the program to execute)
+ may not be a variable, and must be a
+ literal and absolute path
+ name.</para>
+
+ <para>Note that this setting does not
+ directly support shell command
+ lines. If shell command lines are to
+ be used they need to be passed
+ explicitly to a shell implementation
+ of some kind. Example:
+ <literal>ExecStart=/bin/sh -c 'dmesg | tac'</literal></para>
+ </listitem>
</varlistentry>
<varlistentry>
the command in
<varname>ExecStart=</varname>. Multiple
command lines may be concatenated in a
- single directive, by seperating them
+ single directive, by separating them
by semicolons (these semicolons must
- be passed as seperate words). In that
+ be passed as separate words). In that
case, the commands are executed one
after the other,
serially. Alternatively, these
directives may be specified more than
- once whith the same effect. However,
+ once with the same effect. However,
the latter syntax is not recommended
for compatibility with parsers
suitable for XDG
daemon, and may be used for command
lines like the following:
<command>/bin/kill -HUP
- $(MAINPID)</command>.</para></listitem>
+ $MAINPID</command>.</para></listitem>
</varlistentry>
<varlistentry>
configured in this option are run are
terminated according to the
<varname>KillMode=</varname> setting
- (see below). If this option is not
- specified the process is terminated
- right-away when service stop is
- requested. Specifier and environment
- variable substitution is supported
- (including
- <literal>$(MAINPID)</literal>, see
+ (see
+ <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>). If
+ this option is not specified the
+ process is terminated right-away when
+ service stop is requested. Specifier
+ and environment variable substitution
+ is supported (including
+ <literal>$MAINPID</literal>, see
above).</para></listitem>
</varlistentry>
</varlistentry>
<varlistentry>
- <term><varname>TimeoutSec=</varname></term>
+ <term><varname>TimeoutStartSec=</varname></term>
<listitem><para>Configures the time to
- wait for start-up and stop. If a
+ wait for start-up. If a
daemon service does not signal
start-up completion within the
- configured time the service will be
+ configured time, the service will be
considered failed and be shut down
- again. If a service is asked to stop
- but does not terminate in the
- specified time it will be terminated
+ again.
+ Takes a unit-less value in seconds, or a
+ time span value such as "5min
+ 20s". Pass 0 to disable the timeout
+ logic. Defaults to 90s, except when
+ <varname>Type=oneshot</varname> is
+ used in which case the timeout
+ is disabled by default.
+ </para></listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term><varname>TimeoutStopSec=</varname></term>
+ <listitem><para>Configures the time to
+ wait for stop. If a service is asked
+ to stop but does not terminate in the
+ specified time, it will be terminated
forcibly via SIGTERM, and after
another delay of this time with
- SIGKILL. (See
+ SIGKILL (See
<varname>KillMode=</varname>
- below.) Takes a unit-less value in seconds, or a
+ in <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>).
+ Takes a unit-less value in seconds, or a
time span value such as "5min
20s". Pass 0 to disable the timeout
- logic. Defaults to
- 60s.</para></listitem>
+ logic. Defaults to 90s.
+ </para></listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term><varname>TimeoutSec=</varname></term>
+ <listitem><para>A shorthand for configuring
+ both <varname>TimeoutStartSec=</varname>
+ and <varname>TimeoutStopSec=</varname>
+ to the specified value.
+ </para></listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term><varname>WatchdogSec=</varname></term>
+ <listitem><para>Configures the
+ watchdog timeout for a service. This
+ is activated when the start-up is
+ completed. The service must call
+ <citerefentry><refentrytitle>sd_notify</refentrytitle><manvolnum>3</manvolnum></citerefentry>
+ regularly with "WATCHDOG=1" (i.e. the
+ "keep-alive ping"). If the time
+ between two such calls is larger than
+ the configured time then the service
+ is placed in a failure state. By
+ setting <varname>Restart=</varname> to
+ <option>on-failure</option> or
+ <option>always</option> the service
+ will be automatically restarted. The
+ time configured here will be passed to
+ the executed service process in the
+ <varname>WATCHDOG_USEC=</varname>
+ environment variable. This allows
+ daemons to automatically enable the
+ keep-alive pinging logic if watchdog
+ support is enabled for the service. If
+ this option is used
+ <varname>NotifyAccess=</varname> (see
+ below) should be set to open access to
+ the notification socket provided by
+ systemd. If
+ <varname>NotifyAccess=</varname> is
+ not set, it will be implicitly set to
+ <option>main</option>. Defaults to 0,
+ which disables this
+ feature.</para></listitem>
</varlistentry>
<varlistentry>
<term><varname>Restart=</varname></term>
<listitem><para>Configures whether the
- main service process shall be restarted when
- it exists. Takes one of
- <option>once</option>,
- <option>restart-on-success</option> or
- <option>restart-always</option>. If
- set to <option>once</option> (the
- default) the service will not be
- restarted when it exits. If set to
- <option>restart-on-success</option> it
- will be restarted only when it exited
- cleanly, i.e. terminated with an exit
- code of 0. If set to
- <option>restart-always</option> the
- service will be restarted regardless
- whether it exited cleanly or not, or
- got terminated abnormally by a
- signal.</para></listitem>
+ main service process shall be
+ restarted when it exits. Takes one of
+ <option>no</option>,
+ <option>on-success</option>,
+ <option>on-failure</option>,
+ <option>on-abort</option> or
+ <option>always</option>. If set to
+ <option>no</option> (the default) the
+ service will not be restarted when it
+ exits. If set to
+ <option>on-success</option> it will be
+ restarted only when it exited cleanly,
+ i.e. terminated with an exit code of
+ 0. If set to
+ <option>on-failure</option> it will be
+ restarted only when it exited with an
+ exit code not equalling 0, when
+ terminated by a signal (including on
+ core dump), when an operation (such as
+ service reload) times out or when the
+ configured watchdog timeout is
+ triggered. If set to
+ <option>on-abort</option> it will be
+ restarted only if it exits due to
+ reception of an uncaught signal
+ (including on core dump). If set to
+ <option>always</option> the service
+ will be restarted regardless whether
+ it exited cleanly or not, got
+ terminated abnormally by a signal or
+ hit a timeout.</para></listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term><varname>SuccessExitStatus=</varname></term>
+ <listitem><para>Takes a list of exit
+ status definitions that when returned
+ by the main service process will be
+ considered successful termination, in
+ addition to the normal successful exit
+ code 0 and the signals SIGHUP, SIGINT,
+ SIGTERM and SIGPIPE. Exit status
+ definitions can either be numeric exit
+ codes or termination signal names, and
+ are separated by spaces. Example:
+ "<literal>SuccessExitStatus=1 2 8
+ SIGKILL</literal>", ensures that exit
+ codes 1, 2, 8 and the termination
+ signal SIGKILL are considered clean
+ service
+ terminations.</para></listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term><varname>RestartPreventExitStatus=</varname></term>
+ <listitem><para>Takes a list of exit
+ status definitions that when returned
+ by the main service process will
+ prevent automatic service restarts
+ regardless of the restart setting
+ configured with
+ <varname>Restart=</varname>. Exit
+ status definitions can either be
+ numeric exit codes or termination
+ signal names, and are separated by
+ spaces. Defaults to the empty list, so
+ that by default no exit status is
+ excluded from the configured restart
+ logic. Example:
+ "<literal>RestartPreventExitStatus=1 6
+ SIGABRT</literal>", ensures that exit
+ codes 1 and 6 and the termination signal
+ SIGABRT will not result in automatic
+ service restarting.</para></listitem>
</varlistentry>
<varlistentry>
false.</para></listitem>
</varlistentry>
- <varlistentry>
- <term><varname>SysVStartPriority=</varname></term>
- <listitem><para>Set the SysV start
- priority to use to order this service
- in relation to SysV services lacking
- LSB headers. This option is only
- necessary to fix ordering in relation
- to legacy SysV services, that have no
- ordering information encoded in the
- script headers. As such it should only
- be used as temporary compatibility
- option, and not be used in new unit
- files. Almost always it is a better
- choice to add explicit ordering
- directives via
- <varname>After=</varname> or
- <varname>Before=</varname>,
- instead. For more details see
- <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>. If
- used, pass an integer value in the
- range 0-99.</para></listitem>
- </varlistentry>
-
- <varlistentry>
- <term><varname>KillMode=</varname></term>
- <listitem><para>Specifies how
- processes of this service shall be
- killed. One of
- <option>control-group</option>,
- <option>process-group</option>,
- <option>process</option>,
- <option>none</option>.</para>
-
- <para>If set to
- <option>control-group</option> all
- remaining processes in the control
- group of this service will be
- terminated on service stop, after the
- stop command (as configured with
- <varname>ExecStop=</varname>) is
- executed. If set to
- <option>process-group</option> only
- the members of the process group of
- the main service process are
- killed. If set to
- <option>process</option> only the main
- process itself is killed. If set to
- <option>none</option> no process is
- killed. In this case only the stop
- command will be executed on service
- stop, but no process be killed
- otherwise. Processes remaining alive
- after stop are left in their control
- group and the control group continues
- to exist after stop unless it is
- empty. Defaults to
- <option>control-croup</option>.</para>
-
- <para>Processes will first be
- terminated via SIGTERM. If then after
- a delay (configured via the
- <varname>TimeoutSec=</varname> option)
- processes still remain, the
- termination request is repeated with
- the SIGKILL signal. See
- <citerefentry><refentrytitle>kill</refentrytitle><manvolnum>2</manvolnum></citerefentry>
- for more
- information.</para></listitem>
- </varlistentry>
-
<varlistentry>
<term><varname>NonBlocking=</varname></term>
<listitem><para>Set O_NONBLOCK flag
<option>main</option> or
<option>all</option>. If
<option>none</option> no daemon status
- updates are accepted by the service
+ updates are accepted from the service
processes, all status update messages
are ignored. If <option>main</option>
only service updates sent from the
accepted. If <option>all</option> all
services updates from all members of
the service's control group are
- accepted. This option must be set to
+ accepted. This option should be set to
open access to the notification socket
when using
- <varname>Type=notify</varname> (see above).</para></listitem>
+ <varname>Type=notify</varname> or
+ <varname>WatchdogUsec=</varname> (see
+ above). If those options are used but
+ <varname>NotifyAccess=</varname> not
+ configured it will be implicitly set
+ to
+ <option>main</option>.</para></listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term><varname>Sockets=</varname></term>
+ <listitem><para>Specifies the name of
+ the socket units this service shall
+ inherit the sockets from when the
+ service is started. Normally it
+ should not be necessary to use this
+ setting as all sockets whose unit
+ shares the same name as the service
+ (ignoring the different suffix of course)
+ are passed to the spawned
+ process.</para>
+
+ <para>Note that the same socket may be
+ passed to multiple processes at the
+ same time. Also note that a different
+ service may be activated on incoming
+ traffic than inherits the sockets. Or
+ in other words: The
+ <varname>Service=</varname> setting of
+ <filename>.socket</filename> units
+ doesn't have to match the inverse of the
+ <varname>Sockets=</varname> setting of
+ the <filename>.service</filename> it
+ refers to.</para></listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term><varname>StartLimitInterval=</varname></term>
+ <term><varname>StartLimitBurst=</varname></term>
+
+ <listitem><para>Configure service
+ start rate limiting. By default
+ services which are started more often
+ than 5 times within 10s are not
+ permitted to start any more times
+ until the 10s interval ends. With
+ these two options this rate limiting
+ may be modified. Use
+ <varname>StartLimitInterval=</varname>
+ to configure the checking interval
+ (defaults to 10s, set to 0 to disable
+ any kind of rate limiting). Use
+ <varname>StartLimitBurst=</varname> to
+ configure how many starts per interval
+ are allowed (defaults to 5). These
+ configuration options are particularly
+ useful in conjunction with
+ <varname>Restart=</varname>, however
+ apply to all kinds of starts
+ (including manual), not just those
+ triggered by the
+ <varname>Restart=</varname> logic.
+ Note that units which are configured
+ for <varname>Restart=</varname> and
+ which reach the start limit are not
+ attempted to be restarted anymore,
+ however they may still be restarted
+ manually at a later point from which
+ point on the restart logic is again
+ activated. Note that
+ <command>systemctl
+ reset-failed</command> will cause the
+ restart rate counter for a service to
+ be flushed, which is useful if the
+ administrator wants to manually start
+ a service and the start limit
+ interferes with
+ that.</para></listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term><varname>StartLimitAction=</varname></term>
+
+ <listitem><para>Configure the action
+ to take if the rate limit configured
+ with
+ <varname>StartLimitInterval=</varname>
+ and
+ <varname>StartLimitBurst=</varname> is
+ hit. Takes one of
+ <option>none</option>,
+ <option>reboot</option>,
+ <option>reboot-force</option> or
+ <option>reboot-immediate</option>. If
+ <option>none</option> is set,
+ hitting the rate limit will trigger no
+ action besides that the start will not
+ be
+ permitted. <option>reboot</option>
+ causes a reboot following the normal
+ shutdown procedure (i.e. equivalent to
+ <command>systemctl reboot</command>),
+ <option>reboot-force</option> causes
+ an forced reboot which will terminate
+ all processes forcibly but should
+ cause no dirty file systems on reboot
+ (i.e. equivalent to <command>systemctl
+ reboot -f</command>) and
+ <option>reboot-immediate</option>
+ causes immediate execution of the
+ <citerefentry><refentrytitle>reboot</refentrytitle><manvolnum>2</manvolnum></citerefentry>
+ system call, which might result in
+ data loss. Defaults to
+ <option>none</option>.</para></listitem>
+ </varlistentry>
+
+ </variablelist>
+
+ <para>Check
+ <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+ and
+ <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+ for more settings.</para>
+
+ </refsect1>
+
+ <refsect1>
+ <title>Compatibility Options</title>
+
+ <para>The following options are also available in the
+ <literal>[Service]</literal> section, but exist purely
+ for compatibility reasons and should not be used in
+ newly written service files.</para>
+
+ <variablelist>
+ <varlistentry>
+ <term><varname>SysVStartPriority=</varname></term>
+ <listitem><para>Set the SysV start
+ priority to use to order this service
+ in relation to SysV services lacking
+ LSB headers. This option is only
+ necessary to fix ordering in relation
+ to legacy SysV services, that have no
+ ordering information encoded in the
+ script headers. As such it should only
+ be used as temporary compatibility
+ option, and not be used in new unit
+ files. Almost always it is a better
+ choice to add explicit ordering
+ directives via
+ <varname>After=</varname> or
+ <varname>Before=</varname>,
+ instead. For more details see
+ <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>. If
+ used, pass an integer value in the
+ range 0-99.</para></listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term><varname>FsckPassNo=</varname></term>
+ <listitem><para>Set the fsck passno
+ priority to use to order this service
+ in relation to other file system
+ checking services. This option is only
+ necessary to fix ordering in relation
+ to fsck jobs automatically created for
+ all <filename>/etc/fstab</filename>
+ entries with a value in the fs_passno
+ column > 0. As such it should only be
+ used as option for fsck
+ services. Almost always it is a better
+ choice to add explicit ordering
+ directives via
+ <varname>After=</varname> or
+ <varname>Before=</varname>,
+ instead. For more details see
+ <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>. If
+ used, pass an integer value in the
+ same range as
+ <filename>/etc/fstab</filename>'s
+ fs_passno column. See
+ <citerefentry><refentrytitle>fstab</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+ for details.</para></listitem>
</varlistentry>
</variablelist>
<citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
<citerefentry><refentrytitle>systemctl</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
<citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
- <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+ <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+ <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>
</para>
</refsect1>