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 SPDX-License-Identifier: LGPL-2.1+
8 This file is part of elogind.
10 Copyright 2010 Lennart Poettering
13 <refentry id="elogind"
14 xmlns:xi="http://www.w3.org/2001/XInclude">
17 <title>elogind</title>
18 <productname>elogind</productname>
22 <contrib>Developer</contrib>
23 <firstname>Lennart</firstname>
24 <surname>Poettering</surname>
25 <email>lennart@poettering.net</email>
31 <refentrytitle>elogind</refentrytitle>
32 <!-- 0 /// elogind is in section 8
33 <manvolnum>1</manvolnum>
35 <manvolnum>8</manvolnum>
40 <!-- 0 /// Well, elogind isn't that much...
41 <refname>systemd</refname>
42 <refname>init</refname>
43 <refpurpose>systemd system and service manager</refpurpose>
45 <refname>elogind</refname>
46 <refpurpose>Login manager</refpurpose>
51 <!-- 0 /// elogind has a different synopsis of course...
53 <command>/usr/lib/systemd/systemd</command>
54 <arg choice="opt" rep="repeat">OPTIONS</arg>
57 <command>init</command>
58 <arg choice="opt" rep="repeat">OPTIONS</arg>
59 <arg choice="req">COMMAND</arg>
63 <command>elogind</command>
64 <arg choice="opt" rep="repeat">OPTIONS</arg>
65 <arg choice="req">COMMAND</arg>
71 <title>Description</title>
73 <!-- 0 /// elogind does not need that much explanation.
74 <para>systemd is a system and service manager for Linux operating
75 systems. When run as first process on boot (as PID 1), it acts as
76 init system that brings up and maintains userspace
79 <para>For compatibility with SysV, if systemd is called as
80 <command>init</command> and a PID that is not 1, it will execute
81 <command>telinit</command> and pass all command line arguments
82 unmodified. That means <command>init</command> and
83 <command>telinit</command> are mostly equivalent when invoked from
84 normal login sessions. See
85 <citerefentry><refentrytitle>telinit</refentrytitle><manvolnum>8</manvolnum></citerefentry>
86 for more information.</para>
88 <para>When run as a system instance, systemd interprets the
89 configuration file <filename>system.conf</filename> and the files
90 in <filename>system.conf.d</filename> directories; when run as a
91 user instance, systemd interprets the configuration file
92 <filename>user.conf</filename> and the files in
93 <filename>user.conf.d</filename> directories. See
94 <citerefentry><refentrytitle>systemd-system.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>
95 for more information.</para>
97 <para><command>elogind</command> is a system service that
98 manages user logins. It is responsible for:</para>
101 <listitem><para>Keeping track of users and sessions, their
102 processes and their idle state</para></listitem>
104 <listitem><para>Providing PolicyKit-based access for users to
105 operations such as system shutdown or sleep</para></listitem>
107 <listitem><para>Implementing a shutdown/sleep inhibition logic
108 for applications</para></listitem>
110 <listitem><para>Handling of power/sleep hardware
111 keys</para></listitem>
113 <listitem><para>Multi-seat management</para></listitem>
115 <listitem><para>Session switch management</para></listitem>
117 <listitem><para>Device access management for
118 users</para></listitem>
122 <para>User sessions are registered in elogind via the
123 <citerefentry><refentrytitle>pam_elogind</refentrytitle><manvolnum>8</manvolnum></citerefentry>
127 <citerefentry><refentrytitle>logind.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>
128 for information about the configuration of this service.</para>
131 url="http://www.freedesktop.org/wiki/Software/systemd/multiseat">Multi-Seat
132 on Linux</ulink> for an introduction into basic concepts of logind
133 such as users, sessions and seats.</para>
136 url="http://www.freedesktop.org/wiki/Software/systemd/logind">
137 logind D-Bus API Documentation</ulink> for information about the
138 APIs <filename>logind</filename> provides.</para>
140 <para>For more information on the inhibition logic see the <ulink
141 url="http://www.freedesktop.org/wiki/Software/systemd/inhibit">Inhibitor
142 Lock Developer Documentation</ulink>.</para>
147 <title>Options</title>
149 <para>The following options are understood:</para>
152 <!-- 0 /// This list is much shorter with elogind of course.
154 <term><option>--test</option></term>
156 <listitem><para>Determine startup sequence, dump it and exit.
157 This is an option useful for debugging only.</para></listitem>
160 <term><option>--dump-configuration-items</option></term>
162 <listitem><para>Dump understood unit configuration items. This
163 outputs a terse but complete list of configuration items
164 understood in unit definition files.</para></listitem>
167 <term><option>--dump-bus-properties</option></term>
169 <listitem><para>Dump exposed bus properties. This outputs
170 a terse but complete list of properties exposed to dbus.
174 <term><option>--unit=</option></term>
176 <listitem><para>Set default unit to activate on startup. If
177 not specified, defaults to
178 <filename>default.target</filename>.</para></listitem>
181 <term><option>--system</option></term>
182 <term><option>--user</option></term>
184 <listitem><para>For <option>--system</option>, tell systemd to
185 run a system instance, even if the process ID is not 1, i.e.
186 systemd is not run as init process. <option>--user</option>
187 does the opposite, running a user instance even if the process
188 ID is 1. Normally, it should not be necessary to pass these
189 options, as systemd automatically detects the mode it is
190 started in. These options are hence of little use except for
191 debugging. Note that it is not supported booting and
192 maintaining a full system with systemd running in
193 <option>--system</option> mode, but PID not 1. In practice,
194 passing <option>--system</option> explicitly is only useful in
195 conjunction with <option>--test</option>.</para></listitem>
198 <term><option>--dump-core</option></term>
200 <listitem><para>Enable core dumping on crash. This switch has
201 no effect when running as user instance. This setting may also
202 be enabled during boot on the kernel command line via the
203 <varname>systemd.dump_core=</varname> option, see
204 below.</para></listitem>
208 <term><option>--crash-vt=</option><replaceable>VT</replaceable></term>
210 <listitem><para>Switch to a specific virtual console (VT) on
211 crash. Takes a positive integer in the range 1–63, or a
212 boolean argument. If an integer is passed, selects which VT to
213 switch to. If <constant>yes</constant>, the VT kernel messages
214 are written to is selected. If <constant>no</constant>, no VT
215 switch is attempted. This switch has no effect when running as
216 user instance. This setting may also be enabled during boot,
217 on the kernel command line via the
218 <varname>systemd.crash_vt=</varname> option, see
220 <!-- FIXME: there is no crash_vt command line option? -->
221 <!-- 0 /// elogind check_tree.pl can not mangle closed comments.
222 below.</para></listitem>
226 <term><option>--crash-shell</option></term>
228 <listitem><para>Run a shell on crash. This switch has no
229 effect when running as user instance. This setting may also be
230 enabled during boot, on the kernel command line via the
231 <varname>systemd.crash_shell=</varname> option, see
232 below.</para></listitem>
236 <term><option>--crash-reboot</option></term>
238 <listitem><para>Automatically reboot the system on crash. This
239 switch has no effect when running as user instance. This
240 setting may also be enabled during boot, on the kernel command
241 line via the <varname>systemd.crash_reboot=</varname> option,
242 see below.</para></listitem>
246 <term><option>--confirm-spawn</option></term>
248 <listitem><para>Ask for confirmation when spawning processes.
249 This switch has no effect when run as user
250 instance.</para></listitem>
253 <term><option>--show-status=</option></term>
255 <listitem><para>Takes a boolean argument or the special value <constant>auto</constant>. If on, terse unit
256 status information is shown on the console during boot-up and shutdown. If off, no such status information is
257 shown. If set to <constant>auto</constant> behavior is similar to off, except that it is automatically switched
258 to on, as soon as the first unit failure or significant boot delay is encountered. This switch has no effect
259 when invoked as user instance. If specified, overrides both the kernel command line setting
260 <varname>systemd.show_status=</varname> (see below) and the configuration file option
261 <option>ShowStatus=</option>, see
262 <citerefentry><refentrytitle>systemd-system.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
265 <term><option>--log-target=</option></term>
267 <listitem><para>Set log target. Argument must be one of
268 <option>console</option>,
269 <option>journal</option>,
270 <option>kmsg</option>,
271 <option>journal-or-kmsg</option>,
272 <option>null</option>.</para></listitem>
275 <term><option>--log-level=</option></term>
277 <listitem><para>Set log level. As
278 argument this accepts a numerical log
279 level or the well-known <citerefentry project='man-pages'><refentrytitle>syslog</refentrytitle><manvolnum>3</manvolnum></citerefentry>
280 symbolic names (lowercase):
281 <option>emerg</option>,
282 <option>alert</option>,
283 <option>crit</option>,
284 <option>err</option>,
285 <option>warning</option>,
286 <option>notice</option>,
287 <option>info</option>,
288 <option>debug</option>.</para></listitem>
291 <term><option>--log-color=</option></term>
293 <listitem><para>Highlight important log messages. Argument is
294 a boolean value. If the argument is omitted, it defaults to
295 <option>true</option>.</para></listitem>
298 <term><option>--log-location=</option></term>
300 <listitem><para>Include code location in log messages. This is
301 mostly relevant for debugging purposes. Argument is a boolean
302 value. If the argument is omitted it defaults to
303 <option>true</option>.</para></listitem>
306 <term><option>--default-standard-output=</option></term>
307 <term><option>--default-standard-error=</option></term>
309 <listitem><para>Sets the default output or error output for
310 all services and sockets, respectively. That is, controls the
311 default for <option>StandardOutput=</option> and
312 <option>StandardError=</option> (see
313 <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>
314 for details). Takes one of
315 <option>inherit</option>,
316 <option>null</option>,
317 <option>tty</option>,
318 <option>journal</option>,
319 <option>journal+console</option>,
320 <option>syslog</option>,
321 <option>syslog+console</option>,
322 <option>kmsg</option>,
323 <option>kmsg+console</option>. If the
325 <option>--default-standard-output=</option> defaults to
326 <option>journal</option> and
327 <option>--default-standard-error=</option> to
328 <option>inherit</option>.</para></listitem>
332 <term><option>--machine-id=</option></term>
334 <listitem><para>Override the machine-id set on the hard drive,
335 useful for network booting or for containers. May not be set
336 to all zeros.</para></listitem>
340 <term><option>--service-watchdogs=</option></term>
342 <listitem><para>Globally enable/disable all service watchdog timeouts and emergency
343 actions. This setting may also be specified during boot, on the kernel
344 command line via the <varname>systemd.service_watchdogs=</varname>
345 option, see below. Defaults to enabled.</para></listitem>
348 <xi:include href="standard-options.xml" xpointer="help" />
349 <xi:include href="standard-options.xml" xpointer="version" />
352 <term><option>-D</option></term>
353 <term><option>--daemon</option></term>
355 <listitem><para>Daemonize elogind by double-forking
356 into background.</para></listitem>
360 <term><option>-h</option></term>
361 <term><option>--help</option></term>
363 <listitem><para>Show the helpt text and exit.</para></listitem>
369 <!-- 0 /// The whole concpets section is unneeded in elogind
371 <title>Concepts</title>
373 <para>systemd provides a dependency system between various
374 entities called "units" of 11 different types. Units encapsulate
375 various objects that are relevant for system boot-up and
376 maintenance. The majority of units are configured in unit
377 configuration files, whose syntax and basic set of options is
379 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
380 however some are created automatically from other configuration,
381 dynamically from system state or programmatically at runtime.
382 Units may be "active" (meaning started, bound, plugged in, …,
383 depending on the unit type, see below), or "inactive" (meaning
384 stopped, unbound, unplugged, …), as well as in the process of
385 being activated or deactivated, i.e. between the two states (these
386 states are called "activating", "deactivating"). A special
387 "failed" state is available as well, which is very similar to
388 "inactive" and is entered when the service failed in some way
389 (process returned error code on exit, or crashed, an operation
390 timed out, or after too many restarts). If this state is entered,
391 the cause will be logged, for later reference. Note that the
392 various unit types may have a number of additional substates,
393 which are mapped to the five generalized unit states described
396 <para>The following unit types are available:</para>
399 <listitem><para>Service units, which start and control daemons
400 and the processes they consist of. For details, see
401 <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
403 <listitem><para>Socket units, which encapsulate local IPC or
404 network sockets in the system, useful for socket-based
405 activation. For details about socket units, see
406 <citerefentry><refentrytitle>systemd.socket</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
407 for details on socket-based activation and other forms of
409 <citerefentry><refentrytitle>daemon</refentrytitle><manvolnum>7</manvolnum></citerefentry>.</para></listitem>
411 <listitem><para>Target units are useful to group units, or
412 provide well-known synchronization points during boot-up, see
413 <citerefentry><refentrytitle>systemd.target</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
415 <listitem><para>Device units expose kernel devices in systemd
416 and may be used to implement device-based activation. For
418 <citerefentry><refentrytitle>systemd.device</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
420 <listitem><para>Mount units control mount points in the file
421 system, for details see
422 <citerefentry><refentrytitle>systemd.mount</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
424 <listitem><para>Automount units provide automount capabilities,
425 for on-demand mounting of file systems as well as parallelized
427 <citerefentry><refentrytitle>systemd.automount</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
429 <listitem><para>Timer units are useful for triggering activation
430 of other units based on timers. You may find details in
431 <citerefentry><refentrytitle>systemd.timer</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
433 <listitem><para>Swap units are very similar to mount units and
434 encapsulate memory swap partitions or files of the operating
435 system. They are described in
436 <citerefentry><refentrytitle>systemd.swap</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
438 <listitem><para>Path units may be used to activate other
439 services when file system objects change or are modified. See
440 <citerefentry><refentrytitle>systemd.path</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
442 <listitem><para>Slice units may be used to group units which
443 manage system processes (such as service and scope units) in a
444 hierarchical tree for resource management purposes. See
445 <citerefentry><refentrytitle>systemd.slice</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
447 <listitem><para>Scope units are similar to service units, but
448 manage foreign processes instead of starting them as well. See
449 <citerefentry><refentrytitle>systemd.scope</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
453 <para>Units are named as their configuration files. Some units
454 have special semantics. A detailed list is available in
455 <citerefentry><refentrytitle>systemd.special</refentrytitle><manvolnum>7</manvolnum></citerefentry>.</para>
457 <para>systemd knows various kinds of dependencies, including
458 positive and negative requirement dependencies (i.e.
459 <varname>Requires=</varname> and <varname>Conflicts=</varname>) as
460 well as ordering dependencies (<varname>After=</varname> and
461 <varname>Before=</varname>). NB: ordering and requirement
462 dependencies are orthogonal. If only a requirement dependency
463 exists between two units (e.g. <filename>foo.service</filename>
464 requires <filename>bar.service</filename>), but no ordering
465 dependency (e.g. <filename>foo.service</filename> after
466 <filename>bar.service</filename>) and both are requested to start,
467 they will be started in parallel. It is a common pattern that both
468 requirement and ordering dependencies are placed between two
469 units. Also note that the majority of dependencies are implicitly
470 created and maintained by systemd. In most cases, it should be
471 unnecessary to declare additional dependencies manually, however
472 it is possible to do this.</para>
474 <para>Application programs and units (via dependencies) may
475 request state changes of units. In systemd, these requests are
476 encapsulated as 'jobs' and maintained in a job queue. Jobs may
477 succeed or can fail, their execution is ordered based on the
478 ordering dependencies of the units they have been scheduled
481 <para>On boot systemd activates the target unit
482 <filename>default.target</filename> whose job is to activate
483 on-boot services and other on-boot units by pulling them in via
484 dependencies. Usually, the unit name is just an alias (symlink) for
485 either <filename>graphical.target</filename> (for fully-featured
486 boots into the UI) or <filename>multi-user.target</filename> (for
487 limited console-only boots for use in embedded or server
488 environments, or similar; a subset of graphical.target). However,
489 it is at the discretion of the administrator to configure it as an
490 alias to any other target unit. See
491 <citerefentry><refentrytitle>systemd.special</refentrytitle><manvolnum>7</manvolnum></citerefentry>
492 for details about these target units.</para>
494 <para>Processes systemd spawns are placed in individual Linux
495 control groups named after the unit which they belong to in the
496 private systemd hierarchy. (see <ulink
497 url="https://www.kernel.org/doc/Documentation/cgroup-v1/cgroups.txt">cgroups.txt</ulink>
498 for more information about control groups, or short "cgroups").
499 systemd uses this to effectively keep track of processes. Control
500 group information is maintained in the kernel, and is accessible
501 via the file system hierarchy (beneath
502 <filename>/sys/fs/cgroup/systemd/</filename>), or in tools such as
503 <citerefentry project='man-pages'><refentrytitle>systemd-cgls</refentrytitle><manvolnum>1</manvolnum></citerefentry>
505 <citerefentry project='man-pages'><refentrytitle>ps</refentrytitle><manvolnum>1</manvolnum></citerefentry>
506 (<command>ps xawf -eo pid,user,cgroup,args</command> is
507 particularly useful to list all processes and the systemd units
508 they belong to.).</para>
510 <para>systemd is compatible with the SysV init system to a large
511 degree: SysV init scripts are supported and simply read as an
512 alternative (though limited) configuration file format. The SysV
513 <filename>/dev/initctl</filename> interface is provided, and
514 compatibility implementations of the various SysV client tools are
515 available. In addition to that, various established Unix
516 functionality such as <filename>/etc/fstab</filename> or the
517 <filename>utmp</filename> database are supported.</para>
519 <para>systemd has a minimal transaction system: if a unit is
520 requested to start up or shut down it will add it and all its
521 dependencies to a temporary transaction. Then, it will verify if
522 the transaction is consistent (i.e. whether the ordering of all
523 units is cycle-free). If it is not, systemd will try to fix it up,
524 and removes non-essential jobs from the transaction that might
525 remove the loop. Also, systemd tries to suppress non-essential
526 jobs in the transaction that would stop a running service. Finally
527 it is checked whether the jobs of the transaction contradict jobs
528 that have already been queued, and optionally the transaction is
529 aborted then. If all worked out and the transaction is consistent
530 and minimized in its impact it is merged with all already
531 outstanding jobs and added to the run queue. Effectively this
532 means that before executing a requested operation, systemd will
533 verify that it makes sense, fixing it if possible, and only
534 failing if it really cannot work.</para>
536 <para>systemd contains native implementations of various tasks
537 that need to be executed as part of the boot process. For example,
538 it sets the hostname or configures the loopback network device. It
539 also sets up and mounts various API file systems, such as
540 <filename>/sys</filename> or <filename>/proc</filename>.</para>
542 <para>For more information about the concepts and
543 ideas behind systemd, please refer to the
544 <ulink url="http://0pointer.de/blog/projects/systemd.html">Original Design Document</ulink>.</para>
546 <para>Note that some but not all interfaces provided
547 by systemd are covered by the
548 <ulink url="https://www.freedesktop.org/wiki/Software/systemd/InterfaceStabilityPromise">Interface
549 Stability Promise</ulink>.</para>
551 <para>Units may be generated dynamically at boot and system
552 manager reload time, for example based on other configuration
553 files or parameters passed on the kernel command line. For details, see
554 <citerefentry><refentrytitle>systemd.generator</refentrytitle><manvolnum>7</manvolnum></citerefentry>.</para>
556 <para>Systems which invoke systemd in a container or initrd
557 environment should implement the
558 <ulink url="https://www.freedesktop.org/wiki/Software/systemd/ContainerInterface">Container Interface</ulink> or
559 <ulink url="https://www.freedesktop.org/wiki/Software/systemd/InitrdInterface">initrd Interface</ulink>
560 specifications, respectively.</para>
564 <!-- 0 /// The whole directories section is unneeded in elogind
566 <title>Directories</title>
570 <term>System unit directories</term>
572 <listitem><para>The systemd system manager reads unit
573 configuration from various directories. Packages that want to
574 install unit files shall place them in the directory returned
575 by <command>pkg-config systemd
576 --variable=systemdsystemunitdir</command>. Other directories
577 checked are <filename>/usr/local/lib/systemd/system</filename>
578 and <filename>/usr/lib/systemd/system</filename>. User
579 configuration always takes precedence. <command>pkg-config
580 systemd --variable=systemdsystemconfdir</command> returns the
581 path of the system configuration directory. Packages should
582 alter the content of these directories only with the
583 <command>enable</command> and <command>disable</command>
585 <citerefentry><refentrytitle>systemctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>
586 tool. Full list of directories is provided in
587 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
594 <term>User unit directories</term>
596 <listitem><para>Similar rules apply for the user unit
597 directories. However, here the
598 <ulink url="http://standards.freedesktop.org/basedir-spec/basedir-spec-latest.html">XDG
599 Base Directory specification</ulink> is followed to find
600 units. Applications should place their unit files in the
601 directory returned by <command>pkg-config systemd
602 --variable=systemduserunitdir</command>. Global configuration
603 is done in the directory reported by <command>pkg-config
604 systemd --variable=systemduserconfdir</command>. The
605 <command>enable</command> and <command>disable</command>
607 <citerefentry><refentrytitle>systemctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>
608 tool can handle both global (i.e. for all users) and private
609 (for one user) enabling/disabling of units. Full list of
610 directories is provided in
611 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
618 <term>SysV init scripts directory</term>
620 <listitem><para>The location of the SysV init script directory
621 varies between distributions. If systemd cannot find a native
622 unit file for a requested service, it will look for a SysV
623 init script of the same name (with the
624 <filename>.service</filename> suffix
625 removed).</para></listitem>
631 <term>SysV runlevel link farm directory</term>
633 <listitem><para>The location of the SysV runlevel link farm
634 directory varies between distributions. systemd will take the
635 link farm into account when figuring out whether a service
636 shall be enabled. Note that a service unit with a native unit
637 configuration file cannot be started by activating it in the
638 SysV runlevel link farm.</para></listitem>
644 <!-- 0 /// The whole Signals section is unneeded in elogind
646 <title>Signals</title>
650 <term><constant>SIGTERM</constant></term>
652 <listitem><para>Upon receiving this signal the systemd system
653 manager serializes its state, reexecutes itself and
654 deserializes the saved state again. This is mostly equivalent
655 to <command>systemctl daemon-reexec</command>.</para>
657 <para>systemd user managers will start the
658 <filename>exit.target</filename> unit when this signal is
659 received. This is mostly equivalent to <command>systemctl
660 --user start exit.target
661 --job-mode=replace-irreversible</command>.</para></listitem>
665 <term><constant>SIGINT</constant></term>
667 <listitem><para>Upon receiving this signal the systemd system
668 manager will start the
669 <filename>ctrl-alt-del.target</filename> unit. This is mostly
670 equivalent to <command>systemctl start ctrl-alt-del.target
671 --job-mode=replace-irreversible</command>. If this signal is
672 received more than 7 times per 2s, an immediate reboot is
673 triggered. Note that pressing Ctrl-Alt-Del on the console
674 will trigger this signal. Hence, if a reboot is hanging,
675 pressing Ctrl-Alt-Del more than 7 times in 2s is a relatively
676 safe way to trigger an immediate reboot.</para>
678 <para>systemd user managers treat this signal the same way as
679 <constant>SIGTERM</constant>.</para></listitem>
683 <term><constant>SIGWINCH</constant></term>
685 <listitem><para>When this signal is received the systemd
686 system manager will start the
687 <filename>kbrequest.target</filename> unit. This is mostly
688 equivalent to <command>systemctl start
689 kbrequest.target</command>.</para>
691 <para>This signal is ignored by systemd user
692 managers.</para></listitem>
696 <term><constant>SIGPWR</constant></term>
698 <listitem><para>When this signal is received the systemd
699 manager will start the <filename>sigpwr.target</filename>
700 unit. This is mostly equivalent to <command>systemctl start
701 sigpwr.target</command>.</para></listitem>
705 <term><constant>SIGUSR1</constant></term>
707 <listitem><para>When this signal is received the systemd
708 manager will try to reconnect to the D-Bus
709 bus.</para></listitem>
713 <term><constant>SIGUSR2</constant></term>
715 <listitem><para>When this signal is received the systemd
716 manager will log its complete state in human-readable form.
717 The data logged is the same as printed by
718 <command>systemd-analyze dump</command>.</para></listitem>
722 <term><constant>SIGHUP</constant></term>
724 <listitem><para>Reloads the complete daemon configuration.
725 This is mostly equivalent to <command>systemctl
726 daemon-reload</command>.</para></listitem>
730 <term><constant>SIGRTMIN+0</constant></term>
732 <listitem><para>Enters default mode, starts the
733 <filename>default.target</filename> unit. This is mostly
734 equivalent to <command>systemctl isolate
735 default.target</command>.</para></listitem>
739 <term><constant>SIGRTMIN+1</constant></term>
741 <listitem><para>Enters rescue mode, starts the
742 <filename>rescue.target</filename> unit. This is mostly
743 equivalent to <command>systemctl isolate
744 rescue.target</command>.</para></listitem>
748 <term><constant>SIGRTMIN+2</constant></term>
750 <listitem><para>Enters emergency mode, starts the
751 <filename>emergency.service</filename> unit. This is mostly
752 equivalent to <command>systemctl isolate
753 emergency.service</command>.</para></listitem>
757 <term><constant>SIGRTMIN+3</constant></term>
759 <listitem><para>Halts the machine, starts the
760 <filename>halt.target</filename> unit. This is mostly
761 equivalent to <command>systemctl start halt.target
762 --job-mode=replace-irreversible</command>.</para>
767 <term><constant>SIGRTMIN+4</constant></term>
769 <listitem><para>Powers off the machine, starts the
770 <filename>poweroff.target</filename> unit. This is mostly
771 equivalent to <command>systemctl start poweroff.target
772 --job-mode=replace-irreversible</command>.</para>
777 <term><constant>SIGRTMIN+5</constant></term>
779 <listitem><para>Reboots the machine, starts the
780 <filename>reboot.target</filename> unit. This is mostly
781 equivalent to <command>systemctl start reboot.target
782 --job-mode=replace-irreversible</command>.</para>
787 <term><constant>SIGRTMIN+6</constant></term>
789 <listitem><para>Reboots the machine via kexec, starts the
790 <filename>kexec.target</filename> unit. This is mostly
791 equivalent to <command>systemctl start kexec.target
792 --job-mode=replace-irreversible</command>.</para>
797 <term><constant>SIGRTMIN+13</constant></term>
799 <listitem><para>Immediately halts the machine.</para></listitem>
803 <term><constant>SIGRTMIN+14</constant></term>
805 <listitem><para>Immediately powers off the machine.</para></listitem>
809 <term><constant>SIGRTMIN+15</constant></term>
811 <listitem><para>Immediately reboots the machine.</para></listitem>
815 <term><constant>SIGRTMIN+16</constant></term>
817 <listitem><para>Immediately reboots the machine with kexec.</para></listitem>
821 <term><constant>SIGRTMIN+20</constant></term>
823 <listitem><para>Enables display of status messages on the
824 console, as controlled via
825 <varname>systemd.show_status=1</varname> on the kernel command
826 line.</para></listitem>
830 <term><constant>SIGRTMIN+21</constant></term>
832 <listitem><para>Disables display of
833 status messages on the console, as
835 <varname>systemd.show_status=0</varname>
836 on the kernel command
837 line.</para></listitem>
841 <term><constant>SIGRTMIN+22</constant></term>
842 <term><constant>SIGRTMIN+23</constant></term>
844 <listitem><para>Sets the log level to <literal>debug</literal>
845 (or <literal>info</literal> on
846 <constant>SIGRTMIN+23</constant>), as controlled via
847 <varname>systemd.log_level=debug</varname> (or
848 <varname>systemd.log_level=info</varname> on
849 <constant>SIGRTMIN+23</constant>) on the kernel command
850 line.</para></listitem>
854 <term><constant>SIGRTMIN+24</constant></term>
856 <listitem><para>Immediately exits the manager (only available
857 for --user instances).</para></listitem>
861 <term><constant>SIGRTMIN+26</constant></term>
862 <term><constant>SIGRTMIN+27</constant></term>
863 <term><constant>SIGRTMIN+28</constant></term>
865 <listitem><para>Sets the log target to
866 <literal>journal-or-kmsg</literal> (or
867 <literal>console</literal> on
868 <constant>SIGRTMIN+27</constant>, <literal>kmsg</literal> on
869 <constant>SIGRTMIN+28</constant>), as controlled via
870 <varname>systemd.log_target=journal-or-kmsg</varname> (or
871 <varname>systemd.log_target=console</varname> on
872 <constant>SIGRTMIN+27</constant> or
873 <varname>systemd.log_target=kmsg</varname> on
874 <constant>SIGRTMIN+28</constant>) on the kernel command
875 line.</para></listitem>
881 <!-- 0 /// The whole environment section is unneeded in elogind
883 <title>Environment</title>
885 <variablelist class='environment-variables'>
887 <term><varname>$SYSTEMD_LOG_LEVEL</varname></term>
888 <listitem><para>systemd reads the log level from this
889 environment variable. This can be overridden with
890 <option>--log-level=</option>.</para></listitem>
894 <term><varname>$SYSTEMD_LOG_TARGET</varname></term>
895 <listitem><para>systemd reads the log target from this
896 environment variable. This can be overridden with
897 <option>--log-target=</option>.</para></listitem>
901 <term><varname>$SYSTEMD_LOG_COLOR</varname></term>
902 <listitem><para>Controls whether systemd highlights important
903 log messages. This can be overridden with
904 <option>--log-color=</option>.</para></listitem>
908 <term><varname>$SYSTEMD_LOG_LOCATION</varname></term>
909 <listitem><para>Controls whether systemd prints the code
910 location along with log messages. This can be overridden with
911 <option>--log-location=</option>.</para></listitem>
915 <term><varname>$XDG_CONFIG_HOME</varname></term>
916 <term><varname>$XDG_CONFIG_DIRS</varname></term>
917 <term><varname>$XDG_DATA_HOME</varname></term>
918 <term><varname>$XDG_DATA_DIRS</varname></term>
920 <listitem><para>The systemd user manager uses these variables
921 in accordance to the <ulink
922 url="http://standards.freedesktop.org/basedir-spec/basedir-spec-latest.html">XDG
923 Base Directory specification</ulink> to find its
924 configuration.</para></listitem>
928 <term><varname>$SYSTEMD_UNIT_PATH</varname></term>
930 <listitem><para>Controls where systemd looks for unit
931 files.</para></listitem>
935 <term><varname>$SYSTEMD_SYSVINIT_PATH</varname></term>
937 <listitem><para>Controls where systemd looks for SysV init
938 scripts.</para></listitem>
942 <term><varname>$SYSTEMD_SYSVRCND_PATH</varname></term>
944 <listitem><para>Controls where systemd looks for SysV init
945 script runlevel link farms.</para></listitem>
949 <term><varname>$SYSTEMD_COLORS</varname></term>
951 <listitem><para>The value must be a boolean. Controls whether colorized output should be
952 generated. This can be specified to override the decision that <command>systemd</command>
953 makes based on <varname>$TERM</varname> and what the console is connected to.</para>
958 <term><varname>$SYSTEMD_URLIFY</varname></term>
960 <listitem><para>The value must be a boolean. Controls whether clickable links should be generated in the output
961 for terminal emulators supporting this. This can be specified to override the decision that
962 <command>systemd</command> makes based on <varname>$TERM</varname> and other conditions.</para>
967 <term><varname>$LISTEN_PID</varname></term>
968 <term><varname>$LISTEN_FDS</varname></term>
969 <term><varname>$LISTEN_FDNAMES</varname></term>
971 <listitem><para>Set by systemd for supervised processes during
972 socket-based activation. See
973 <citerefentry><refentrytitle>sd_listen_fds</refentrytitle><manvolnum>3</manvolnum></citerefentry>
974 for more information.</para></listitem>
978 <term><varname>$NOTIFY_SOCKET</varname></term>
980 <listitem><para>Set by systemd for supervised processes for
981 status and start-up completion notification. See
982 <citerefentry><refentrytitle>sd_notify</refentrytitle><manvolnum>3</manvolnum></citerefentry>
983 for more information.</para></listitem>
989 <!-- 0 /// The whole kernel section is unneeded in elogind
991 <title>Kernel Command Line</title>
993 <para>When run as system instance systemd parses a number of
994 kernel command line arguments<footnote><para>If run inside a Linux
995 container these arguments may be passed as command line arguments
996 to systemd itself, next to any of the command line options listed
997 in the Options section above. If run outside of Linux containers,
998 these arguments are parsed from <filename>/proc/cmdline</filename>
999 instead.</para></footnote>:</para>
1001 <variablelist class='kernel-commandline-options'>
1003 <term><varname>systemd.unit=</varname></term>
1004 <term><varname>rd.systemd.unit=</varname></term>
1006 <listitem><para>Overrides the unit to activate on boot.
1007 Defaults to <filename>default.target</filename>. This may be
1008 used to temporarily boot into a different boot unit, for
1009 example <filename>rescue.target</filename> or
1010 <filename>emergency.service</filename>. See
1011 <citerefentry><refentrytitle>systemd.special</refentrytitle><manvolnum>7</manvolnum></citerefentry>
1012 for details about these units. The option prefixed with
1013 <literal>rd.</literal> is honored only in the initial RAM disk
1014 (initrd), while the one that is not prefixed only in the main
1015 system.</para></listitem>
1019 <term><varname>systemd.dump_core</varname></term>
1021 <listitem><para>Takes a boolean argument or enables the option if specified
1022 without an argument. If enabled, the systemd manager (PID 1) dumps core when
1023 it crashes. Otherwise, no core dump is created. Defaults to enabled.</para>
1028 <term><varname>systemd.crash_chvt</varname></term>
1030 <listitem><para>Takes a positive integer, or a boolean argument. Can be also
1031 specified without an argument, with the same effect as a positive boolean. If
1032 a positive integer (in the range 1–63) is specified, the system manager (PID
1033 1) will activate the specified virtual terminal (VT) when it
1034 crashes. Defaults to disabled, meaning that no such switch is attempted. If
1035 set to enabled, the VT the kernel messages are written to is selected.
1040 <term><varname>systemd.crash_shell</varname></term>
1042 <listitem><para>Takes a boolean argument or enables the option if specified
1043 without an argument. If enabled, the system manager (PID 1) spawns a shell
1044 when it crashes, after a 10s delay. Otherwise, no shell is spawned. Defaults
1045 to disabled, for security reasons, as the shell is not protected by password
1046 authentication.</para></listitem>
1050 <term><varname>systemd.crash_reboot</varname></term>
1052 <listitem><para>Takes a boolean argument or enables the option if specified
1053 without an argument. If enabled, the system manager (PID 1) will reboot the
1054 machine automatically when it crashes, after a 10s delay. Otherwise, the
1055 system will hang indefinitely. Defaults to disabled, in order to avoid a
1056 reboot loop. If combined with <varname>systemd.crash_shell</varname>, the
1057 system is rebooted after the shell exits.</para></listitem>
1061 <term><varname>systemd.confirm_spawn</varname></term>
1063 <listitem><para>Takes a boolean argument or a path to the virtual console
1064 where the confirmation messages should be emitted. Can be also specified
1065 without an argument, with the same effect as a positive boolean. If enabled,
1066 the system manager (PID 1) asks for confirmation when spawning processes
1067 using <option>/dev/console</option>. If a path or a console name (such as
1068 <literal>ttyS0</literal>) is provided, the virtual console pointed to by this
1069 path or described by the give name will be used instead. Defaults to disabled.
1074 <term><varname>systemd.service_watchdogs=</varname></term>
1076 <listitem><para>Takes a boolean argument. If disabled, all service runtime
1077 watchdogs (<option>WatchdogSec=</option>) and emergency actions (e.g.
1078 <option>OnFailure=</option> or <option>StartLimitAction=</option>) are
1079 ignored by the system manager (PID 1); see
1080 <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
1081 Defaults to enabled, i.e. watchdogs and failure actions are processed
1082 normally. The hardware watchdog is not affected by this
1083 option.</para></listitem>
1087 <term><varname>systemd.show_status</varname></term>
1089 <listitem><para>Takes a boolean argument or the constant
1090 <constant>auto</constant>. Can be also specified without an argument, with
1091 the same effect as a positive boolean. If enabled, the systemd manager (PID
1092 1) shows terse service status updates on the console during bootup.
1093 <constant>auto</constant> behaves like <option>false</option> until a unit
1094 fails or there is a significant delay in boot. Defaults to enabled, unless
1095 <option>quiet</option> is passed as kernel command line option, in which case
1096 it defaults to <constant>auto</constant>. If specified overrides the system
1097 manager configuration file option <option>ShowStatus=</option>, see
1098 <citerefentry><refentrytitle>systemd-system.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
1099 However, the process command line option <option>--show-status=</option>
1100 takes precedence over both this kernel command line option and the
1101 configuration file option.</para></listitem>
1105 <term><varname>systemd.log_target=</varname></term>
1106 <term><varname>systemd.log_level=</varname></term>
1107 <term><varname>systemd.log_location=</varname></term>
1108 <term><varname>systemd.log_color</varname></term>
1110 <listitem><para>Controls log output, with the same effect as the
1111 <varname>$SYSTEMD_LOG_TARGET</varname>,
1112 <varname>$SYSTEMD_LOG_LEVEL</varname>,
1113 <varname>$SYSTEMD_LOG_LOCATION</varname>,
1114 <varname>$SYSTEMD_LOG_COLOR</varname> environment variables described above.
1115 <varname>systemd.log_color</varname> can be specified without an argument,
1116 with the same effect as a positive boolean.</para></listitem>
1120 <term><varname>systemd.default_standard_output=</varname></term>
1121 <term><varname>systemd.default_standard_error=</varname></term>
1122 <listitem><para>Controls default standard output and error
1123 output for services, with the same effect as the
1124 <option>--default-standard-output=</option> and
1125 <option>--default-standard-error=</option> command line
1126 arguments described above, respectively.</para></listitem>
1130 <term><varname>systemd.setenv=</varname></term>
1132 <listitem><para>Takes a string argument in the form
1133 VARIABLE=VALUE. May be used to set default environment
1134 variables to add to forked child processes. May be used more
1135 than once to set multiple variables.</para></listitem>
1139 <term><varname>systemd.machine_id=</varname></term>
1141 <listitem><para>Takes a 32 character hex value to be
1142 used for setting the machine-id. Intended mostly for
1143 network booting where the same machine-id is desired
1144 for every boot.</para></listitem>
1148 <term><varname>systemd.unified_cgroup_hierarchy</varname></term>
1150 <listitem><para>When specified without an argument or with a true argument,
1151 enables the usage of
1152 <ulink url="https://www.kernel.org/doc/Documentation/cgroup-v2.txt">unified cgroup hierarchy</ulink>
1153 (a.k.a. cgroups-v2). When specified with a false argument, fall back to
1154 hybrid or full legacy cgroup hierarchy.</para>
1156 <para>If this option is not specified, the default behaviour is determined
1157 during compilation (the <option>--with-default-hierarchy=</option>
1158 option). If the kernel does not support unified cgroup hierarchy, the legacy
1159 hierarchy will be used even if this option is specified.</para>
1164 <term><varname>systemd.legacy_systemd_cgroup_controller</varname></term>
1166 <listitem><para>Takes effect if the full unified cgroup hierarchy is not used
1167 (see previous option). When specified without an argument or with a true
1168 argument, disables the use of "hybrid" cgroup hierarchy (i.e. a cgroups-v2
1169 tree used for systemd, and
1170 <ulink url="https://www.kernel.org/doc/Documentation/cgroup-v1/">legacy
1171 cgroup hierarchy</ulink>, a.k.a. cgroups-v1, for other controllers), and
1172 forces a full "legacy" mode. When specified with a false argument, enables
1173 the use of "hybrid" hierarchy.</para>
1175 <para>If this option is not specified, the default behaviour is determined
1176 during compilation (the <option>--with-default-hierarchy=</option>
1177 option). If the kernel does not support unified cgroup hierarchy, the legacy
1178 hierarchy will be used even if this option is specified.</para>
1183 <term><varname>quiet</varname></term>
1185 <listitem><para>Turn off status output at boot, much like
1186 <varname>systemd.show_status=false</varname> would. Note that
1187 this option is also read by the kernel itself and disables
1188 kernel log output. Passing this option hence turns off the
1189 usual output from both the system manager and the kernel.
1194 <term><varname>debug</varname></term>
1196 <listitem><para>Turn on debugging output. This is equivalent
1197 to <varname>systemd.log_level=debug</varname>. Note that this
1198 option is also read by the kernel itself and enables kernel
1199 debug output. Passing this option hence turns on the debug
1200 output from both the system manager and the
1201 kernel.</para></listitem>
1205 <term><varname>emergency</varname></term>
1206 <term><varname>rd.emergency</varname></term>
1207 <term><varname>-b</varname></term>
1209 <listitem><para>Boot into emergency mode. This is equivalent
1210 to <varname>systemd.unit=emergency.target</varname> or
1211 <varname>rd.systemd.unit=emergency.target</varname>, respectively, and
1212 provided for compatibility reasons and to be easier to type.</para></listitem>
1216 <term><varname>rescue</varname></term>
1217 <term><varname>rd.rescue</varname></term>
1218 <term><varname>single</varname></term>
1219 <term><varname>s</varname></term>
1220 <term><varname>S</varname></term>
1221 <term><varname>1</varname></term>
1223 <listitem><para>Boot into rescue mode. This is equivalent to
1224 <varname>systemd.unit=rescue.target</varname> or
1225 <varname>rd.systemd.unit=rescue.target</varname>, respectively, and
1226 provided for compatibility reasons and to be easier to type.</para></listitem>
1230 <term><varname>2</varname></term>
1231 <term><varname>3</varname></term>
1232 <term><varname>4</varname></term>
1233 <term><varname>5</varname></term>
1235 <listitem><para>Boot into the specified legacy SysV runlevel.
1236 These are equivalent to
1237 <varname>systemd.unit=runlevel2.target</varname>,
1238 <varname>systemd.unit=runlevel3.target</varname>,
1239 <varname>systemd.unit=runlevel4.target</varname>, and
1240 <varname>systemd.unit=runlevel5.target</varname>,
1241 respectively, and provided for compatibility reasons and to be
1242 easier to type.</para></listitem>
1246 <term><varname>locale.LANG=</varname></term>
1247 <term><varname>locale.LANGUAGE=</varname></term>
1248 <term><varname>locale.LC_CTYPE=</varname></term>
1249 <term><varname>locale.LC_NUMERIC=</varname></term>
1250 <term><varname>locale.LC_TIME=</varname></term>
1251 <term><varname>locale.LC_COLLATE=</varname></term>
1252 <term><varname>locale.LC_MONETARY=</varname></term>
1253 <term><varname>locale.LC_MESSAGES=</varname></term>
1254 <term><varname>locale.LC_PAPER=</varname></term>
1255 <term><varname>locale.LC_NAME=</varname></term>
1256 <term><varname>locale.LC_ADDRESS=</varname></term>
1257 <term><varname>locale.LC_TELEPHONE=</varname></term>
1258 <term><varname>locale.LC_MEASUREMENT=</varname></term>
1259 <term><varname>locale.LC_IDENTIFICATION=</varname></term>
1261 <listitem><para>Set the system locale to use. This overrides
1262 the settings in <filename>/etc/locale.conf</filename>. For
1263 more information, see
1264 <citerefentry project='man-pages'><refentrytitle>locale.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>
1266 <citerefentry project='man-pages'><refentrytitle>locale</refentrytitle><manvolnum>7</manvolnum></citerefentry>.
1271 <para>For other kernel command line parameters understood by
1272 components of the core OS, please refer to
1273 <citerefentry><refentrytitle>kernel-command-line</refentrytitle><manvolnum>7</manvolnum></citerefentry>.</para>
1277 <!-- 0 /// The whole sockets section is unneeded in elogind
1279 <title>Sockets and FIFOs</title>
1283 <term><filename>/run/systemd/notify</filename></term>
1285 <listitem><para>Daemon status notification socket. This is an
1286 <constant>AF_UNIX</constant> datagram socket and is used to
1287 implement the daemon notification logic as implemented by
1288 <citerefentry><refentrytitle>sd_notify</refentrytitle><manvolnum>3</manvolnum></citerefentry>.</para></listitem>
1293 <term><filename>/run/systemd/private</filename></term>
1295 <listitem><para>Used internally as communication channel
1297 <citerefentry><refentrytitle>systemctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>
1298 and the systemd process. This is an
1299 <constant>AF_UNIX</constant> stream socket. This interface is
1300 private to systemd and should not be used in external
1301 projects.</para></listitem>
1305 <term><filename>/dev/initctl</filename></term>
1307 <listitem><para>Limited compatibility support for the SysV
1308 client interface, as implemented by the
1309 <filename>systemd-initctl.service</filename> unit. This is a
1310 named pipe in the file system. This interface is obsolete and
1311 should not be used in new applications.</para></listitem>
1318 <title>See Also</title>
1320 <!-- 0 /// elogind does not have that much to see also...
1321 The <ulink url="https://www.freedesktop.org/wiki/Software/systemd/">systemd Homepage</ulink>,
1322 <citerefentry><refentrytitle>systemd-system.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1323 <citerefentry project='man-pages'><refentrytitle>locale.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1324 <citerefentry><refentrytitle>systemctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1325 <citerefentry><refentrytitle>journalctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1326 <citerefentry><refentrytitle>systemd-notify</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1327 <citerefentry><refentrytitle>daemon</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
1328 <citerefentry><refentrytitle>sd-daemon</refentrytitle><manvolnum>3</manvolnum></citerefentry>,
1329 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1330 <citerefentry><refentrytitle>systemd.special</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1331 <citerefentry project='die-net'><refentrytitle>pkg-config</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1332 <citerefentry><refentrytitle>kernel-command-line</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
1333 <citerefentry project='man-pages'><refentrytitle>bootup</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
1334 <citerefentry><refentrytitle>systemd.directives</refentrytitle><manvolnum>7</manvolnum></citerefentry>
1336 The <ulink url="https://github.com/elogind/elogind/">elogind Homepage</ulink>,
1337 <citerefentry><refentrytitle>loginctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1338 <citerefentry><refentrytitle>logind.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1339 <citerefentry><refentrytitle>pam_elogind</refentrytitle><manvolnum>8</manvolnum></citerefentry>