1 <?xml version='
1.0'
?> <!--*- Mode: nxml; nxml-child-indent: 2; indent-tabs-mode: nil -*-->
2 <!DOCTYPE refentry PUBLIC
"-//OASIS//DTD DocBook XML V4.2//EN"
3 "http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
6 This file is part of systemd.
8 Copyright 2010 Lennart Poettering
10 systemd is free software; you can redistribute it and/or modify it
11 under the terms of the GNU Lesser General Public License as published by
12 the Free Software Foundation; either version 2.1 of the License, or
13 (at your option) any later version.
15 systemd is distributed in the hope that it will be useful, but
16 WITHOUT ANY WARRANTY; without even the implied warranty of
17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
18 Lesser General Public License for more details.
20 You should have received a copy of the GNU Lesser General Public License
21 along with systemd; If not, see <http://www.gnu.org/licenses/>.
24 <refentry id=
"systemd"
25 xmlns:
xi=
"http://www.w3.org/2001/XInclude">
28 <title>systemd
</title>
29 <productname>systemd
</productname>
33 <contrib>Developer
</contrib>
34 <firstname>Lennart
</firstname>
35 <surname>Poettering
</surname>
36 <email>lennart@poettering.net
</email>
42 <refentrytitle>systemd
</refentrytitle>
43 <manvolnum>1</manvolnum>
47 <refname>systemd
</refname>
48 <refname>init
</refname>
49 <refpurpose>systemd system and service manager
</refpurpose>
54 <command>systemd
</command>
55 <arg choice=
"opt" rep=
"repeat">OPTIONS
</arg>
58 <command>init
</command>
59 <arg choice=
"opt" rep=
"repeat">OPTIONS
</arg>
60 <arg choice=
"req">COMMAND
</arg>
65 <title>Description
</title>
67 <para>systemd is a system and service manager for Linux operating
68 systems. When run as first process on boot (as PID
1), it acts as
69 init system that brings up and maintains userspace
72 <para>For compatibility with SysV, if systemd is called as
73 <command>init
</command> and a PID that is not
1, it will execute
74 <command>telinit
</command> and pass all command line arguments
75 unmodified. That means
<command>init
</command> and
76 <command>telinit
</command> are mostly equivalent when invoked from
77 normal login sessions. See
78 <citerefentry><refentrytitle>telinit
</refentrytitle><manvolnum>8</manvolnum></citerefentry>
79 for more information.
</para>
81 <para>When run as a system instance, systemd interprets the
82 configuration file
<filename>system.conf
</filename> and the files
83 in
<filename>system.conf.d
</filename> directories; when run as a
84 user instance, systemd interprets the configuration file
85 <filename>user.conf
</filename> and the files in
86 <filename>user.conf.d
</filename> directories. See
87 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>
88 for more information.
</para>
92 <title>Options
</title>
94 <para>The following options are understood:
</para>
98 <term><option>--test
</option></term>
100 <listitem><para>Determine startup sequence, dump it and exit.
101 This is an option useful for debugging only.
</para></listitem>
104 <term><option>--dump-configuration-items
</option></term>
106 <listitem><para>Dump understood unit configuration items. This
107 outputs a terse but complete list of configuration items
108 understood in unit definition files.
</para></listitem>
111 <term><option>--unit=
</option></term>
113 <listitem><para>Set default unit to activate on startup. If
114 not specified, defaults to
115 <filename>default.target
</filename>.
</para></listitem>
118 <term><option>--system
</option></term>
119 <term><option>--user
</option></term>
121 <listitem><para>For
<option>--system
</option>, tell systemd to
122 run a system instance, even if the process ID is not
1, i.e.
123 systemd is not run as init process.
<option>--user
</option>
124 does the opposite, running a user instance even if the process
125 ID is
1. Normally, it should not be necessary to pass these
126 options, as systemd automatically detects the mode it is
127 started in. These options are hence of little use except for
128 debugging. Note that it is not supported booting and
129 maintaining a full system with systemd running in
130 <option>--system
</option> mode, but PID not
1. In practice,
131 passing
<option>--system
</option> explicitly is only useful in
132 conjunction with
<option>--test
</option>.
</para></listitem>
135 <term><option>--dump-core
</option></term>
137 <listitem><para>Enable core dumping on crash. This switch has
138 no effect when running as user instance. This setting may also
139 be enabled during boot on the kernel command line via the
140 <varname>systemd.dump_core=
</varname> option, see
141 below.
</para></listitem>
145 <term><option>--crash-vt=
</option><replaceable>VT
</replaceable></term>
147 <listitem><para>Switch to a specific virtual console (VT) on
148 crash. Takes a positive integer in the range
1–
63, or a
149 boolean argument. If an integer is passed, selects which VT to
150 switch to. If
<constant>yes
</constant>, the VT kernel messages
151 are written to is selected. If
<constant>no
</constant>, no VT
152 switch is attempted. This switch has no effect when running as
153 user instance. This setting may also be enabled during boot,
154 on the kernel command line via the
155 <varname>systemd.crash_vt=
</varname> option, see
156 <!-- FIXME: there is no crash_vt command line option? -->
157 below.
</para></listitem>
161 <term><option>--crash-shell
</option></term>
163 <listitem><para>Run a shell on crash. This switch has no
164 effect when running as user instance. This setting may also be
165 enabled during boot, on the kernel command line via the
166 <varname>systemd.crash_shell=
</varname> option, see
167 below.
</para></listitem>
171 <term><option>--crash-reboot
</option></term>
173 <listitem><para>Automatically reboot the system on crash. This
174 switch has no effect when running as user instance. This
175 setting may also be enabled during boot, on the kernel command
176 line via the
<varname>systemd.crash_reboot=
</varname> option,
177 see below.
</para></listitem>
181 <term><option>--confirm-spawn
</option></term>
183 <listitem><para>Ask for confirmation when spawning processes.
184 This switch has no effect when run as user
185 instance.
</para></listitem>
188 <term><option>--show-status=
</option></term>
190 <listitem><para>Takes a boolean argument or the special value
<constant>auto
</constant>. If on, terse unit
191 status information is shown on the console during boot-up and shutdown. If off, no such status information is
192 shown. If set to
<constant>auto
</constant> behavior is similar to off, except that it is automatically switched
193 to on, as soon as the first unit failure or significant boot delay is encountered. This switch has no effect
194 when invoked as user instance. If specified, overrides both the kernel command line setting
195 <varname>systemd.show_status=
</varname> (see below) and the configuration file option
196 <option>ShowStatus=
</option>, see
197 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
200 <term><option>--log-target=
</option></term>
202 <listitem><para>Set log target. Argument must be one of
203 <option>console
</option>,
204 <option>journal
</option>,
205 <option>kmsg
</option>,
206 <option>journal-or-kmsg
</option>,
207 <option>null
</option>.
</para></listitem>
210 <term><option>--log-level=
</option></term>
212 <listitem><para>Set log level. As
213 argument this accepts a numerical log
214 level or the well-known
<citerefentry project='man-pages'
><refentrytitle>syslog
</refentrytitle><manvolnum>3</manvolnum></citerefentry>
215 symbolic names (lowercase):
216 <option>emerg
</option>,
217 <option>alert
</option>,
218 <option>crit
</option>,
219 <option>err
</option>,
220 <option>warning
</option>,
221 <option>notice
</option>,
222 <option>info
</option>,
223 <option>debug
</option>.
</para></listitem>
226 <term><option>--log-color=
</option></term>
228 <listitem><para>Highlight important log messages. Argument is
229 a boolean value. If the argument is omitted, it defaults to
230 <option>true
</option>.
</para></listitem>
233 <term><option>--log-location=
</option></term>
235 <listitem><para>Include code location in log messages. This is
236 mostly relevant for debugging purposes. Argument is a boolean
237 value. If the argument is omitted it defaults to
238 <option>true
</option>.
</para></listitem>
241 <term><option>--default-standard-output=
</option></term>
242 <term><option>--default-standard-error=
</option></term>
244 <listitem><para>Sets the default output or error output for
245 all services and sockets, respectively. That is, controls the
246 default for
<option>StandardOutput=
</option> and
247 <option>StandardError=
</option> (see
248 <citerefentry><refentrytitle>systemd.exec
</refentrytitle><manvolnum>5</manvolnum></citerefentry>
249 for details). Takes one of
250 <option>inherit
</option>,
251 <option>null
</option>,
252 <option>tty
</option>,
253 <option>journal
</option>,
254 <option>journal+console
</option>,
255 <option>syslog
</option>,
256 <option>syslog+console
</option>,
257 <option>kmsg
</option>,
258 <option>kmsg+console
</option>. If the
260 <option>--default-standard-output=
</option> defaults to
261 <option>journal
</option> and
262 <option>--default-standard-error=
</option> to
263 <option>inherit
</option>.
</para></listitem>
267 <term><option>--machine-id=
</option></term>
269 <listitem><para>Override the machine-id set on the hard drive,
270 useful for network booting or for containers. May not be set
271 to all zeros.
</para></listitem>
274 <xi:include href=
"standard-options.xml" xpointer=
"help" />
275 <xi:include href=
"standard-options.xml" xpointer=
"version" />
280 <title>Concepts
</title>
282 <para>systemd provides a dependency system between various
283 entities called
"units" of
11 different types. Units encapsulate
284 various objects that are relevant for system boot-up and
285 maintenance. The majority of units are configured in unit
286 configuration files, whose syntax and basic set of options is
288 <citerefentry><refentrytitle>systemd.unit
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
289 however some are created automatically from other configuration,
290 dynamically from system state or programmatically at runtime.
291 Units may be
"active" (meaning started, bound, plugged in, …,
292 depending on the unit type, see below), or
"inactive" (meaning
293 stopped, unbound, unplugged, …), as well as in the process of
294 being activated or deactivated, i.e. between the two states (these
295 states are called
"activating",
"deactivating"). A special
296 "failed" state is available as well, which is very similar to
297 "inactive" and is entered when the service failed in some way
298 (process returned error code on exit, or crashed, or an operation
299 timed out). If this state is entered, the cause will be logged,
300 for later reference. Note that the various unit types may have a
301 number of additional substates, which are mapped to the five
302 generalized unit states described here.
</para>
304 <para>The following unit types are available:
</para>
307 <listitem><para>Service units, which start and control daemons
308 and the processes they consist of. For details, see
309 <citerefentry><refentrytitle>systemd.service
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
311 <listitem><para>Socket units, which encapsulate local IPC or
312 network sockets in the system, useful for socket-based
313 activation. For details about socket units, see
314 <citerefentry><refentrytitle>systemd.socket
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
315 for details on socket-based activation and other forms of
317 <citerefentry><refentrytitle>daemon
</refentrytitle><manvolnum>7</manvolnum></citerefentry>.
</para></listitem>
319 <listitem><para>Target units are useful to group units, or
320 provide well-known synchronization points during boot-up, see
321 <citerefentry><refentrytitle>systemd.target
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
323 <listitem><para>Device units expose kernel devices in systemd
324 and may be used to implement device-based activation. For
326 <citerefentry><refentrytitle>systemd.device
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
328 <listitem><para>Mount units control mount points in the file
329 system, for details see
330 <citerefentry><refentrytitle>systemd.mount
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
332 <listitem><para>Automount units provide automount capabilities,
333 for on-demand mounting of file systems as well as parallelized
335 <citerefentry><refentrytitle>systemd.automount
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
337 <listitem><para>Timer units are useful for triggering activation
338 of other units based on timers. You may find details in
339 <citerefentry><refentrytitle>systemd.timer
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
341 <listitem><para>Swap units are very similar to mount units and
342 encapsulate memory swap partitions or files of the operating
343 system. They are described in
344 <citerefentry><refentrytitle>systemd.swap
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
346 <listitem><para>Path units may be used to activate other
347 services when file system objects change or are modified. See
348 <citerefentry><refentrytitle>systemd.path
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
350 <listitem><para>Slice units may be used to group units which
351 manage system processes (such as service and scope units) in a
352 hierarchical tree for resource management purposes. See
353 <citerefentry><refentrytitle>systemd.slice
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
355 <listitem><para>Scope units are similar to service units, but
356 manage foreign processes instead of starting them as well. See
357 <citerefentry><refentrytitle>systemd.scope
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para></listitem>
361 <para>Units are named as their configuration files. Some units
362 have special semantics. A detailed list is available in
363 <citerefentry><refentrytitle>systemd.special
</refentrytitle><manvolnum>7</manvolnum></citerefentry>.
</para>
365 <para>systemd knows various kinds of dependencies, including
366 positive and negative requirement dependencies (i.e.
367 <varname>Requires=
</varname> and
<varname>Conflicts=
</varname>) as
368 well as ordering dependencies (
<varname>After=
</varname> and
369 <varname>Before=
</varname>). NB: ordering and requirement
370 dependencies are orthogonal. If only a requirement dependency
371 exists between two units (e.g.
<filename>foo.service
</filename>
372 requires
<filename>bar.service
</filename>), but no ordering
373 dependency (e.g.
<filename>foo.service
</filename> after
374 <filename>bar.service
</filename>) and both are requested to start,
375 they will be started in parallel. It is a common pattern that both
376 requirement and ordering dependencies are placed between two
377 units. Also note that the majority of dependencies are implicitly
378 created and maintained by systemd. In most cases, it should be
379 unnecessary to declare additional dependencies manually, however
380 it is possible to do this.
</para>
382 <para>Application programs and units (via dependencies) may
383 request state changes of units. In systemd, these requests are
384 encapsulated as 'jobs' and maintained in a job queue. Jobs may
385 succeed or can fail, their execution is ordered based on the
386 ordering dependencies of the units they have been scheduled
389 <para>On boot systemd activates the target unit
390 <filename>default.target
</filename> whose job is to activate
391 on-boot services and other on-boot units by pulling them in via
392 dependencies. Usually, the unit name is just an alias (symlink) for
393 either
<filename>graphical.target
</filename> (for fully-featured
394 boots into the UI) or
<filename>multi-user.target
</filename> (for
395 limited console-only boots for use in embedded or server
396 environments, or similar; a subset of graphical.target). However,
397 it is at the discretion of the administrator to configure it as an
398 alias to any other target unit. See
399 <citerefentry><refentrytitle>systemd.special
</refentrytitle><manvolnum>7</manvolnum></citerefentry>
400 for details about these target units.
</para>
402 <para>Processes systemd spawns are placed in individual Linux
403 control groups named after the unit which they belong to in the
404 private systemd hierarchy. (see
<ulink
405 url=
"https://www.kernel.org/doc/Documentation/cgroups/cgroups.txt">cgroups.txt
</ulink>
406 for more information about control groups, or short
"cgroups").
407 systemd uses this to effectively keep track of processes. Control
408 group information is maintained in the kernel, and is accessible
409 via the file system hierarchy (beneath
410 <filename>/sys/fs/cgroup/systemd/
</filename>), or in tools such as
411 <citerefentry project='man-pages'
><refentrytitle>systemd-cgls
</refentrytitle><manvolnum>1</manvolnum></citerefentry>
413 <citerefentry project='man-pages'
><refentrytitle>ps
</refentrytitle><manvolnum>1</manvolnum></citerefentry>
414 (
<command>ps xawf -eo pid,user,cgroup,args
</command> is
415 particularly useful to list all processes and the systemd units
416 they belong to.).
</para>
418 <para>systemd is compatible with the SysV init system to a large
419 degree: SysV init scripts are supported and simply read as an
420 alternative (though limited) configuration file format. The SysV
421 <filename>/dev/initctl
</filename> interface is provided, and
422 compatibility implementations of the various SysV client tools are
423 available. In addition to that, various established Unix
424 functionality such as
<filename>/etc/fstab
</filename> or the
425 <filename>utmp
</filename> database are supported.
</para>
427 <para>systemd has a minimal transaction system: if a unit is
428 requested to start up or shut down it will add it and all its
429 dependencies to a temporary transaction. Then, it will verify if
430 the transaction is consistent (i.e. whether the ordering of all
431 units is cycle-free). If it is not, systemd will try to fix it up,
432 and removes non-essential jobs from the transaction that might
433 remove the loop. Also, systemd tries to suppress non-essential
434 jobs in the transaction that would stop a running service. Finally
435 it is checked whether the jobs of the transaction contradict jobs
436 that have already been queued, and optionally the transaction is
437 aborted then. If all worked out and the transaction is consistent
438 and minimized in its impact it is merged with all already
439 outstanding jobs and added to the run queue. Effectively this
440 means that before executing a requested operation, systemd will
441 verify that it makes sense, fixing it if possible, and only
442 failing if it really cannot work.
</para>
444 <para>Systemd contains native implementations of various tasks
445 that need to be executed as part of the boot process. For example,
446 it sets the hostname or configures the loopback network device. It
447 also sets up and mounts various API file systems, such as
448 <filename>/sys
</filename> or
<filename>/proc
</filename>.
</para>
450 <para>For more information about the concepts and
451 ideas behind systemd, please refer to the
452 <ulink url=
"http://0pointer.de/blog/projects/systemd.html">Original Design Document
</ulink>.
</para>
454 <para>Note that some but not all interfaces provided
455 by systemd are covered by the
456 <ulink url=
"https://www.freedesktop.org/wiki/Software/systemd/InterfaceStabilityPromise">Interface
457 Stability Promise
</ulink>.
</para>
459 <para>Units may be generated dynamically at boot and system
460 manager reload time, for example based on other configuration
461 files or parameters passed on the kernel command line. For details, see
462 <citerefentry><refentrytitle>systemd.generator
</refentrytitle><manvolnum>7</manvolnum></citerefentry>.
</para>
464 <para>Systems which invoke systemd in a container or initrd
465 environment should implement the
466 <ulink url=
"https://www.freedesktop.org/wiki/Software/systemd/ContainerInterface">Container Interface
</ulink> or
467 <ulink url=
"https://www.freedesktop.org/wiki/Software/systemd/InitrdInterface">initrd Interface
</ulink>
468 specifications, respectively.
</para>
472 <title>Directories
</title>
476 <term>System unit directories
</term>
478 <listitem><para>The systemd system manager reads unit
479 configuration from various directories. Packages that want to
480 install unit files shall place them in the directory returned
481 by
<command>pkg-config systemd
482 --variable=systemdsystemunitdir
</command>. Other directories
483 checked are
<filename>/usr/local/lib/systemd/system
</filename>
484 and
<filename>/usr/lib/systemd/system
</filename>. User
485 configuration always takes precedence.
<command>pkg-config
486 systemd --variable=systemdsystemconfdir
</command> returns the
487 path of the system configuration directory. Packages should
488 alter the content of these directories only with the
489 <command>enable
</command> and
<command>disable
</command>
491 <citerefentry><refentrytitle>systemctl
</refentrytitle><manvolnum>1</manvolnum></citerefentry>
492 tool. Full list of directories is provided in
493 <citerefentry><refentrytitle>systemd.unit
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
500 <term>User unit directories
</term>
502 <listitem><para>Similar rules apply for the user unit
503 directories. However, here the
504 <ulink url=
"http://standards.freedesktop.org/basedir-spec/basedir-spec-latest.html">XDG
505 Base Directory specification
</ulink> is followed to find
506 units. Applications should place their unit files in the
507 directory returned by
<command>pkg-config systemd
508 --variable=systemduserunitdir
</command>. Global configuration
509 is done in the directory reported by
<command>pkg-config
510 systemd --variable=systemduserconfdir
</command>. The
511 <command>enable
</command> and
<command>disable
</command>
513 <citerefentry><refentrytitle>systemctl
</refentrytitle><manvolnum>1</manvolnum></citerefentry>
514 tool can handle both global (i.e. for all users) and private
515 (for one user) enabling/disabling of units. Full list of
516 directories is provided in
517 <citerefentry><refentrytitle>systemd.unit
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
524 <term>SysV init scripts directory
</term>
526 <listitem><para>The location of the SysV init script directory
527 varies between distributions. If systemd cannot find a native
528 unit file for a requested service, it will look for a SysV
529 init script of the same name (with the
530 <filename>.service
</filename> suffix
531 removed).
</para></listitem>
537 <term>SysV runlevel link farm directory
</term>
539 <listitem><para>The location of the SysV runlevel link farm
540 directory varies between distributions. systemd will take the
541 link farm into account when figuring out whether a service
542 shall be enabled. Note that a service unit with a native unit
543 configuration file cannot be started by activating it in the
544 SysV runlevel link farm.
</para></listitem>
550 <title>Signals
</title>
554 <term><constant>SIGTERM
</constant></term>
556 <listitem><para>Upon receiving this signal the systemd system
557 manager serializes its state, reexecutes itself and
558 deserializes the saved state again. This is mostly equivalent
559 to
<command>systemctl daemon-reexec
</command>.
</para>
561 <para>systemd user managers will start the
562 <filename>exit.target
</filename> unit when this signal is
563 received. This is mostly equivalent to
<command>systemctl
564 --user start exit.target
</command>.
</para></listitem>
568 <term><constant>SIGINT
</constant></term>
570 <listitem><para>Upon receiving this signal the systemd system
571 manager will start the
572 <filename>ctrl-alt-del.target
</filename> unit. This is mostly
573 equivalent to
<command>systemctl start
574 ctl-alt-del.target
</command>. If this signal is received more
575 than
7 times per
2s, an immediate reboot is triggered.
576 Note that pressing Ctrl-Alt-Del on the console will trigger
577 this signal. Hence, if a reboot is hanging, pressing
578 Ctrl-Alt-Del more than
7 times in
2s is a relatively safe way
579 to trigger an immediate reboot.
</para>
581 <para>systemd user managers treat this signal the same way as
582 <constant>SIGTERM
</constant>.
</para></listitem>
586 <term><constant>SIGWINCH
</constant></term>
588 <listitem><para>When this signal is received the systemd
589 system manager will start the
590 <filename>kbrequest.target
</filename> unit. This is mostly
591 equivalent to
<command>systemctl start
592 kbrequest.target
</command>.
</para>
594 <para>This signal is ignored by systemd user
595 managers.
</para></listitem>
599 <term><constant>SIGPWR
</constant></term>
601 <listitem><para>When this signal is received the systemd
602 manager will start the
<filename>sigpwr.target
</filename>
603 unit. This is mostly equivalent to
<command>systemctl start
604 sigpwr.target
</command>.
</para></listitem>
608 <term><constant>SIGUSR1
</constant></term>
610 <listitem><para>When this signal is received the systemd
611 manager will try to reconnect to the D-Bus
612 bus.
</para></listitem>
616 <term><constant>SIGUSR2
</constant></term>
618 <listitem><para>When this signal is received the systemd
619 manager will log its complete state in human-readable form.
620 The data logged is the same as printed by
621 <command>systemd-analyze dump
</command>.
</para></listitem>
625 <term><constant>SIGHUP
</constant></term>
627 <listitem><para>Reloads the complete daemon configuration.
628 This is mostly equivalent to
<command>systemctl
629 daemon-reload
</command>.
</para></listitem>
633 <term><constant>SIGRTMIN+
0</constant></term>
635 <listitem><para>Enters default mode, starts the
636 <filename>default.target
</filename> unit. This is mostly
637 equivalent to
<command>systemctl start
638 default.target
</command>.
</para></listitem>
642 <term><constant>SIGRTMIN+
1</constant></term>
644 <listitem><para>Enters rescue mode, starts the
645 <filename>rescue.target
</filename> unit. This is mostly
646 equivalent to
<command>systemctl isolate
647 rescue.target
</command>.
</para></listitem>
651 <term><constant>SIGRTMIN+
2</constant></term>
653 <listitem><para>Enters emergency mode, starts the
654 <filename>emergency.service
</filename> unit. This is mostly
655 equivalent to
<command>systemctl isolate
656 emergency.service
</command>.
</para></listitem>
660 <term><constant>SIGRTMIN+
3</constant></term>
662 <listitem><para>Halts the machine, starts the
663 <filename>halt.target
</filename> unit. This is mostly
664 equivalent to
<command>systemctl start
665 halt.target
</command>.
</para></listitem>
669 <term><constant>SIGRTMIN+
4</constant></term>
671 <listitem><para>Powers off the machine, starts the
672 <filename>poweroff.target
</filename> unit. This is mostly
673 equivalent to
<command>systemctl start
674 poweroff.target
</command>.
</para></listitem>
678 <term><constant>SIGRTMIN+
5</constant></term>
680 <listitem><para>Reboots the machine, starts the
681 <filename>reboot.target
</filename> unit. This is mostly
682 equivalent to
<command>systemctl start
683 reboot.target
</command>.
</para></listitem>
687 <term><constant>SIGRTMIN+
6</constant></term>
689 <listitem><para>Reboots the machine via kexec, starts the
690 <filename>kexec.target
</filename> unit. This is mostly
691 equivalent to
<command>systemctl start
692 kexec.target
</command>.
</para></listitem>
696 <term><constant>SIGRTMIN+
13</constant></term>
698 <listitem><para>Immediately halts the machine.
</para></listitem>
702 <term><constant>SIGRTMIN+
14</constant></term>
704 <listitem><para>Immediately powers off the machine.
</para></listitem>
708 <term><constant>SIGRTMIN+
15</constant></term>
710 <listitem><para>Immediately reboots the machine.
</para></listitem>
714 <term><constant>SIGRTMIN+
16</constant></term>
716 <listitem><para>Immediately reboots the machine with kexec.
</para></listitem>
720 <term><constant>SIGRTMIN+
20</constant></term>
722 <listitem><para>Enables display of status messages on the
723 console, as controlled via
724 <varname>systemd.show_status=
1</varname> on the kernel command
725 line.
</para></listitem>
729 <term><constant>SIGRTMIN+
21</constant></term>
731 <listitem><para>Disables display of
732 status messages on the console, as
734 <varname>systemd.show_status=
0</varname>
735 on the kernel command
736 line.
</para></listitem>
740 <term><constant>SIGRTMIN+
22</constant></term>
741 <term><constant>SIGRTMIN+
23</constant></term>
743 <listitem><para>Sets the log level to
<literal>debug
</literal>
744 (or
<literal>info
</literal> on
745 <constant>SIGRTMIN+
23</constant>), as controlled via
746 <varname>systemd.log_level=debug
</varname> (or
747 <varname>systemd.log_level=info
</varname> on
748 <constant>SIGRTMIN+
23</constant>) on the kernel command
749 line.
</para></listitem>
753 <term><constant>SIGRTMIN+
24</constant></term>
755 <listitem><para>Immediately exits the manager (only available
756 for --user instances).
</para></listitem>
760 <term><constant>SIGRTMIN+
26</constant></term>
761 <term><constant>SIGRTMIN+
27</constant></term>
762 <term><constant>SIGRTMIN+
28</constant></term>
764 <listitem><para>Sets the log target to
765 <literal>journal-or-kmsg
</literal> (or
766 <literal>console
</literal> on
767 <constant>SIGRTMIN+
27</constant>,
<literal>kmsg
</literal> on
768 <constant>SIGRTMIN+
28</constant>), as controlled via
769 <varname>systemd.log_target=journal-or-kmsg
</varname> (or
770 <varname>systemd.log_target=console
</varname> on
771 <constant>SIGRTMIN+
27</constant> or
772 <varname>systemd.log_target=kmsg
</varname> on
773 <constant>SIGRTMIN+
28</constant>) on the kernel command
774 line.
</para></listitem>
780 <title>Environment
</title>
782 <variablelist class='environment-variables'
>
784 <term><varname>$SYSTEMD_LOG_LEVEL
</varname></term>
785 <listitem><para>systemd reads the log level from this
786 environment variable. This can be overridden with
787 <option>--log-level=
</option>.
</para></listitem>
791 <term><varname>$SYSTEMD_LOG_TARGET
</varname></term>
792 <listitem><para>systemd reads the log target from this
793 environment variable. This can be overridden with
794 <option>--log-target=
</option>.
</para></listitem>
798 <term><varname>$SYSTEMD_LOG_COLOR
</varname></term>
799 <listitem><para>Controls whether systemd highlights important
800 log messages. This can be overridden with
801 <option>--log-color=
</option>.
</para></listitem>
805 <term><varname>$SYSTEMD_LOG_LOCATION
</varname></term>
806 <listitem><para>Controls whether systemd prints the code
807 location along with log messages. This can be overridden with
808 <option>--log-location=
</option>.
</para></listitem>
812 <term><varname>$XDG_CONFIG_HOME
</varname></term>
813 <term><varname>$XDG_CONFIG_DIRS
</varname></term>
814 <term><varname>$XDG_DATA_HOME
</varname></term>
815 <term><varname>$XDG_DATA_DIRS
</varname></term>
817 <listitem><para>The systemd user manager uses these variables
818 in accordance to the
<ulink
819 url=
"http://standards.freedesktop.org/basedir-spec/basedir-spec-latest.html">XDG
820 Base Directory specification
</ulink> to find its
821 configuration.
</para></listitem>
825 <term><varname>$SYSTEMD_UNIT_PATH
</varname></term>
827 <listitem><para>Controls where systemd looks for unit
828 files.
</para></listitem>
832 <term><varname>$SYSTEMD_SYSVINIT_PATH
</varname></term>
834 <listitem><para>Controls where systemd looks for SysV init
835 scripts.
</para></listitem>
839 <term><varname>$SYSTEMD_SYSVRCND_PATH
</varname></term>
841 <listitem><para>Controls where systemd looks for SysV init
842 script runlevel link farms.
</para></listitem>
846 <term><varname>$SYSTEMD_COLORS
</varname></term>
848 <listitem><para>The value must be a boolean. Controls whether colorized output should be
849 generated. This can be specified to override the decision that
<command>systemd
</command>
850 makes based on
<varname>$TERM
</varname> and what the console is connected to.
</para>
855 <term><varname>$LISTEN_PID
</varname></term>
856 <term><varname>$LISTEN_FDS
</varname></term>
857 <term><varname>$LISTEN_FDNAMES
</varname></term>
859 <listitem><para>Set by systemd for supervised processes during
860 socket-based activation. See
861 <citerefentry><refentrytitle>sd_listen_fds
</refentrytitle><manvolnum>3</manvolnum></citerefentry>
862 for more information.
</para></listitem>
866 <term><varname>$NOTIFY_SOCKET
</varname></term>
868 <listitem><para>Set by systemd for supervised processes for
869 status and start-up completion notification. See
870 <citerefentry><refentrytitle>sd_notify
</refentrytitle><manvolnum>3</manvolnum></citerefentry>
871 for more information.
</para></listitem>
877 <title>Kernel Command Line
</title>
879 <para>When run as system instance systemd parses a number of
880 kernel command line arguments
<footnote><para>If run inside a Linux
881 container these arguments may be passed as command line arguments
882 to systemd itself, next to any of the command line options listed
883 in the Options section above. If run outside of Linux containers,
884 these arguments are parsed from
<filename>/proc/cmdline
</filename>
885 instead.
</para></footnote>:
</para>
887 <variablelist class='kernel-commandline-options'
>
889 <term><varname>systemd.unit=
</varname></term>
890 <term><varname>rd.systemd.unit=
</varname></term>
892 <listitem><para>Overrides the unit to activate on boot.
893 Defaults to
<filename>default.target
</filename>. This may be
894 used to temporarily boot into a different boot unit, for
895 example
<filename>rescue.target
</filename> or
896 <filename>emergency.service
</filename>. See
897 <citerefentry><refentrytitle>systemd.special
</refentrytitle><manvolnum>7</manvolnum></citerefentry>
898 for details about these units. The option prefixed with
899 <literal>rd.
</literal> is honored only in the initial RAM disk
900 (initrd), while the one that is not prefixed only in the main
901 system.
</para></listitem>
905 <term><varname>systemd.dump_core
</varname></term>
907 <listitem><para>Takes a boolean argument or enables the option if specified
908 without an argument. If enabled, the systemd manager (PID
1) dumps core when
909 it crashes. Otherwise, no core dump is created. Defaults to enabled.
</para>
914 <term><varname>systemd.crash_chvt
</varname></term>
916 <listitem><para>Takes a positive integer, or a boolean argument. Can be also
917 specified without an argument, with the same effect as a positive boolean. If
918 a positive integer (in the range
1–
63) is specified, the system manager (PID
919 1) will activate the specified virtual terminal (VT) when it
920 crashes. Defaults to disabled, meaning that no such switch is attempted. If
921 set to enabled, the VT the kernel messages are written to is selected.
926 <term><varname>systemd.crash_shell
</varname></term>
928 <listitem><para>Takes a boolean argument or enables the option if specified
929 without an argument. If enabled, the system manager (PID
1) spawns a shell
930 when it crashes, after a
10s delay. Otherwise, no shell is spawned. Defaults
931 to disabled, for security reasons, as the shell is not protected by password
932 authentication.
</para></listitem>
936 <term><varname>systemd.crash_reboot
</varname></term>
938 <listitem><para>Takes a boolean argument or enables the option if specified
939 without an argument. If enabled, the system manager (PID
1) will reboot the
940 machine automatically when it crashes, after a
10s delay. Otherwise, the
941 system will hang indefinitely. Defaults to disabled, in order to avoid a
942 reboot loop. If combined with
<varname>systemd.crash_shell
</varname>, the
943 system is rebooted after the shell exits.
</para></listitem>
947 <term><varname>systemd.confirm_spawn
</varname></term>
949 <listitem><para>Takes a boolean argument or a path to the virtual console
950 where the confirmation messages should be emitted. Can be also specified
951 without an argument, with the same effect as a positive boolean. If enabled,
952 the system manager (PID
1) asks for confirmation when spawning processes
953 using
<option>/dev/console
</option>. If a path or a console name (such as
954 <literal>ttyS0
</literal>) is provided, the virtual console pointed to by this
955 path or described by the give name will be used instead. Defaults to disabled.
960 <term><varname>systemd.show_status
</varname></term>
962 <listitem><para>Takes a boolean argument or the constant
963 <constant>auto
</constant>. Can be also specified without an argument, with
964 the same effect as a positive boolean. If enabled, the systemd manager (PID
965 1) shows terse service status updates on the console during bootup.
966 <constant>auto
</constant> behaves like
<option>false
</option> until a unit
967 fails or there is a significant delay in boot. Defaults to enabled, unless
968 <option>quiet
</option> is passed as kernel command line option, in which case
969 it defaults to
<constant>auto
</constant>. If specified overrides the system
970 manager configuration file option
<option>ShowStatus=
</option>, see
971 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
972 However, the process command line option
<option>--show-status=
</option>
973 takes precedence over both this kernel command line option and the
974 configuration file option.
</para></listitem>
978 <term><varname>systemd.log_target=
</varname></term>
979 <term><varname>systemd.log_level=
</varname></term>
980 <term><varname>systemd.log_location=
</varname></term>
981 <term><varname>systemd.log_color
</varname></term>
983 <listitem><para>Controls log output, with the same effect as the
984 <varname>$SYSTEMD_LOG_TARGET
</varname>,
985 <varname>$SYSTEMD_LOG_LEVEL
</varname>,
986 <varname>$SYSTEMD_LOG_LOCATION
</varname>,
987 <varname>$SYSTEMD_LOG_COLOR
</varname> environment variables described above.
988 <varname>systemd.log_color
</varname> can be specified without an argument,
989 with the same effect as a positive boolean.
</para></listitem>
993 <term><varname>systemd.default_standard_output=
</varname></term>
994 <term><varname>systemd.default_standard_error=
</varname></term>
995 <listitem><para>Controls default standard output and error
996 output for services, with the same effect as the
997 <option>--default-standard-output=
</option> and
998 <option>--default-standard-error=
</option> command line
999 arguments described above, respectively.
</para></listitem>
1003 <term><varname>systemd.setenv=
</varname></term>
1005 <listitem><para>Takes a string argument in the form
1006 VARIABLE=VALUE. May be used to set default environment
1007 variables to add to forked child processes. May be used more
1008 than once to set multiple variables.
</para></listitem>
1012 <term><varname>systemd.machine_id=
</varname></term>
1014 <listitem><para>Takes a
32 character hex value to be
1015 used for setting the machine-id. Intended mostly for
1016 network booting where the same machine-id is desired
1017 for every boot.
</para></listitem>
1021 <term><varname>systemd.unified_cgroup_hierarchy
</varname></term>
1023 <listitem><para>When specified without an argument or with a true argument,
1024 enables the usage of
1025 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">unified cgroup hierarchy
</ulink>
1026 (a.k.a. cgroups-v2). When specified with a false argument, fall back to
1027 hybrid or full legacy cgroup hierarchy.
</para>
1029 <para>If this option is not specified, the default behaviour is determined
1030 during compilation (the
<option>--with-default-hierarchy=
</option>
1031 option). If the kernel does not support unified cgroup hierarchy, the legacy
1032 hierarchy will be used even if this option is specified.
</para>
1037 <term><varname>systemd.legacy_systemd_cgroup_controller
</varname></term>
1039 <listitem><para>Takes effect if the full unified cgroup hierarchy is not used
1040 (see previous option). When specified without an argument or with a true
1041 argument, disables the use of
"hybrid" cgroup hierarchy (i.e. a cgroups-v2
1042 tree used for systemd, and
1043 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/">legacy
1044 cgroup hierarchy
</ulink>, a.k.a. cgroups-v1, for other controllers), and
1045 forces a full
"legacy" mode. When specified with a false argument, enables
1046 the use of
"hybrid" hierarchy.
</para>
1048 <para>If this option is not specified, the default behaviour is determined
1049 during compilation (the
<option>--with-default-hierarchy=
</option>
1050 option). If the kernel does not support unified cgroup hierarchy, the legacy
1051 hierarchy will be used even if this option is specified.
</para>
1056 <term><varname>quiet
</varname></term>
1058 <listitem><para>Turn off status output at boot, much like
1059 <varname>systemd.show_status=false
</varname> would. Note that
1060 this option is also read by the kernel itself and disables
1061 kernel log output. Passing this option hence turns off the
1062 usual output from both the system manager and the kernel.
1067 <term><varname>debug
</varname></term>
1069 <listitem><para>Turn on debugging output. This is equivalent
1070 to
<varname>systemd.log_level=debug
</varname>. Note that this
1071 option is also read by the kernel itself and enables kernel
1072 debug output. Passing this option hence turns on the debug
1073 output from both the system manager and the
1074 kernel.
</para></listitem>
1078 <term><varname>emergency
</varname></term>
1079 <term><varname>rd.emergency
</varname></term>
1080 <term><varname>-b
</varname></term>
1082 <listitem><para>Boot into emergency mode. This is equivalent
1083 to
<varname>systemd.unit=emergency.target
</varname> or
1084 <varname>rd.systemd.unit=emergency.target
</varname>, respectively, and
1085 provided for compatibility reasons and to be easier to type.
</para></listitem>
1089 <term><varname>rescue
</varname></term>
1090 <term><varname>rd.rescue
</varname></term>
1091 <term><varname>single
</varname></term>
1092 <term><varname>s
</varname></term>
1093 <term><varname>S
</varname></term>
1094 <term><varname>1</varname></term>
1096 <listitem><para>Boot into rescue mode. This is equivalent to
1097 <varname>systemd.unit=rescue.target
</varname> or
1098 <varname>rd.systemd.unit=rescue.target
</varname>, respectively, and
1099 provided for compatibility reasons and to be easier to type.
</para></listitem>
1103 <term><varname>2</varname></term>
1104 <term><varname>3</varname></term>
1105 <term><varname>4</varname></term>
1106 <term><varname>5</varname></term>
1108 <listitem><para>Boot into the specified legacy SysV runlevel.
1109 These are equivalent to
1110 <varname>systemd.unit=runlevel2.target
</varname>,
1111 <varname>systemd.unit=runlevel3.target
</varname>,
1112 <varname>systemd.unit=runlevel4.target
</varname>, and
1113 <varname>systemd.unit=runlevel5.target
</varname>,
1114 respectively, and provided for compatibility reasons and to be
1115 easier to type.
</para></listitem>
1119 <term><varname>locale.LANG=
</varname></term>
1120 <term><varname>locale.LANGUAGE=
</varname></term>
1121 <term><varname>locale.LC_CTYPE=
</varname></term>
1122 <term><varname>locale.LC_NUMERIC=
</varname></term>
1123 <term><varname>locale.LC_TIME=
</varname></term>
1124 <term><varname>locale.LC_COLLATE=
</varname></term>
1125 <term><varname>locale.LC_MONETARY=
</varname></term>
1126 <term><varname>locale.LC_MESSAGES=
</varname></term>
1127 <term><varname>locale.LC_PAPER=
</varname></term>
1128 <term><varname>locale.LC_NAME=
</varname></term>
1129 <term><varname>locale.LC_ADDRESS=
</varname></term>
1130 <term><varname>locale.LC_TELEPHONE=
</varname></term>
1131 <term><varname>locale.LC_MEASUREMENT=
</varname></term>
1132 <term><varname>locale.LC_IDENTIFICATION=
</varname></term>
1134 <listitem><para>Set the system locale to use. This overrides
1135 the settings in
<filename>/etc/locale.conf
</filename>. For
1136 more information, see
1137 <citerefentry project='man-pages'
><refentrytitle>locale.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>
1139 <citerefentry project='man-pages'
><refentrytitle>locale
</refentrytitle><manvolnum>7</manvolnum></citerefentry>.
1144 <para>For other kernel command line parameters understood by
1145 components of the core OS, please refer to
1146 <citerefentry><refentrytitle>kernel-command-line
</refentrytitle><manvolnum>7</manvolnum></citerefentry>.
</para>
1150 <title>Sockets and FIFOs
</title>
1154 <term><filename>/run/systemd/notify
</filename></term>
1156 <listitem><para>Daemon status notification socket. This is an
1157 <constant>AF_UNIX
</constant> datagram socket and is used to
1158 implement the daemon notification logic as implemented by
1159 <citerefentry><refentrytitle>sd_notify
</refentrytitle><manvolnum>3</manvolnum></citerefentry>.
</para></listitem>
1164 <term><filename>/run/systemd/private
</filename></term>
1166 <listitem><para>Used internally as communication channel
1168 <citerefentry><refentrytitle>systemctl
</refentrytitle><manvolnum>1</manvolnum></citerefentry>
1169 and the systemd process. This is an
1170 <constant>AF_UNIX
</constant> stream socket. This interface is
1171 private to systemd and should not be used in external
1172 projects.
</para></listitem>
1176 <term><filename>/dev/initctl
</filename></term>
1178 <listitem><para>Limited compatibility support for the SysV
1179 client interface, as implemented by the
1180 <filename>systemd-initctl.service
</filename> unit. This is a
1181 named pipe in the file system. This interface is obsolete and
1182 should not be used in new applications.
</para></listitem>
1188 <title>See Also
</title>
1190 The
<ulink url=
"https://www.freedesktop.org/wiki/Software/systemd/">systemd Homepage
</ulink>,
1191 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1192 <citerefentry project='man-pages'
><refentrytitle>locale.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1193 <citerefentry><refentrytitle>systemctl
</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1194 <citerefentry><refentrytitle>journalctl
</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1195 <citerefentry><refentrytitle>systemd-notify
</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1196 <citerefentry><refentrytitle>daemon
</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
1197 <citerefentry><refentrytitle>sd-daemon
</refentrytitle><manvolnum>3</manvolnum></citerefentry>,
1198 <citerefentry><refentrytitle>systemd.unit
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1199 <citerefentry><refentrytitle>systemd.special
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1200 <citerefentry project='die-net'
><refentrytitle>pkg-config
</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1201 <citerefentry><refentrytitle>kernel-command-line
</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
1202 <citerefentry project='man-pages'
><refentrytitle>bootup
</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
1203 <citerefentry><refentrytitle>systemd.directives
</refentrytitle><manvolnum>7</manvolnum></citerefentry>