1 <?xml version='1.0'?> <!--*-nxml-*-->
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="systemctl">
27 <title>systemctl</title>
28 <productname>systemd</productname>
32 <contrib>Developer</contrib>
33 <firstname>Lennart</firstname>
34 <surname>Poettering</surname>
35 <email>lennart@poettering.net</email>
41 <refentrytitle>systemctl</refentrytitle>
42 <manvolnum>1</manvolnum>
46 <refname>systemctl</refname>
47 <refpurpose>Control the systemd system and service manager</refpurpose>
52 <command>systemctl</command>
53 <arg choice="opt" rep="repeat">OPTIONS</arg>
54 <arg choice="plain">COMMAND</arg>
55 <arg choice="opt" rep="repeat">NAME</arg>
60 <title>Description</title>
62 <para><command>systemctl</command> may be used to
63 introspect and control the state of the
64 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>
65 system and service manager.</para>
69 <title>Options</title>
71 <para>The following options are understood:</para>
75 <term><option>-h</option></term>
76 <term><option>--help</option></term>
78 <listitem><para>Prints a short help
79 text and exits.</para></listitem>
83 <term><option>--version</option></term>
86 <para>Prints a short version string and exits.</para>
91 <term><option>-t</option></term>
92 <term><option>--type=</option></term>
95 <para>The argument should be a unit type name such as
96 <option>service</option> and <option>socket</option>, or a
97 unit load state such as <option>loaded</option> and
98 <option>masked</option>.
101 <para>If the argument is a unit type, when listing units,
102 limit display to certain unit types. If not specified units
103 of all types will be shown.</para>
105 <para>If the argument is a unit load state, when listing
106 units, limit display to certain unit types. If not specified
107 units of in all load states will be shown.</para>
109 <para>As a special case, if the argument is
110 <option>help</option>, a list of allowed values will be
111 printed and the program will exit.</para>
116 <term><option>-p</option></term>
117 <term><option>--property=</option></term>
120 <para>When showing unit/job/manager properties, limit
121 display to certain properties as specified as argument. If
122 not specified all set properties are shown. The argument
123 should be a comma-seperated list of property names, such as
124 <literal>MainPID</literal>. If specified more than once all
125 properties with the specified names are shown.</para>
130 <term><option>-a</option></term>
131 <term><option>--all</option></term>
134 <para>When listing units, show all units, regardless of
135 their state, including inactive units. When showing
136 unit/job/manager properties, show all properties regardless
137 whether they are set or not.</para>
142 <term><option>--failed</option></term>
145 <para>When listing units, show only failed units. Do not
146 confuse with <option>--fail</option>.</para>
151 <term><option>--full</option></term>
154 <para>Do not ellipsize unit names, cgroup members, and
155 truncate unit descriptions in the output of
156 <command>list-units</command> and
157 <command>list-jobs</command>.</para>
162 <term><option>--fail</option></term>
165 <para>If the requested operation conflicts with a pending
166 unfinished job, fail the command. If this is not specified
167 the requested operation will replace the pending job, if
168 necessary. Do not confuse with
169 <option>--failed</option>.</para>
174 <term><option>--irreversible</option></term>
177 <para>Mark this transaction's jobs as irreversible. This prevents
178 future conflicting transactions from replacing these jobs.
179 The jobs can still be cancelled using the <command>cancel</command>
185 <term><option>--ignore-dependencies</option></term>
188 <para>When enqueuing a new job ignore all its dependencies
189 and execute it immediately. If passed no required units of
190 the unit passed will be pulled in, and no ordering
191 dependencies will be honored. This is mostly a debugging and
192 rescue tool for the administrator and should not be used by
198 <term><option>-i</option></term>
199 <term><option>--ignore-inhibitors</option></term>
202 <para>When system shutdown or a sleep state is requested,
203 ignore inhibitor locks. Applications can establish inhibitor
204 locks to avoid that certain important operations (such as CD
205 burning or suchlike) are interrupted by system shutdown or a
206 sleep state. Any user may take these locks and privileged
207 users may override these locks. If any locks are taken,
208 shutdown and sleep state requests will normally fail
209 (regardless if privileged or not) and list of active locks
210 is printed. However if <option>--ignore-inhibitors</option>
211 is specified the locks are ignored and not printed, and the
212 operation attempted anyway, possibly requiring additional
218 <term><option>-q</option></term>
219 <term><option>--quiet</option></term>
222 <para>Suppress output to standard output in
223 <command>snapshot</command>,
224 <command>is-active</command>,
225 <command>is-failed</command>,
226 <command>enable</command> and
227 <command>disable</command>.</para>
232 <term><option>--no-block</option></term>
235 <para>Do not synchronously wait for the requested operation
236 to finish. If this is not specified the job will be
237 verified, enqueued and <command>systemctl</command> will
238 wait until it is completed. By passing this argument it is
239 only verified and enqueued.</para>
244 <term><option>--no-legend</option></term>
247 <para>Do not print a legend, i.e. the column headers and
248 the footer with hints.</para>
253 <term><option>--no-pager</option></term>
256 <para>Do not pipe output into a pager.</para>
261 <term><option>--system</option></term>
264 <para>Talk to the systemd system manager. (Default)</para>
269 <term><option>--user</option></term>
272 <para>Talk to the systemd manager of the calling
278 <term><option>--no-wall</option></term>
281 <para>Don't send wall message before halt, power-off,
287 <term><option>--global</option></term>
290 <para>When used with <command>enable</command> and
291 <command>disable</command>, operate on the global user
292 configuration directory, thus enabling or disabling a unit
293 file globally for all future logins of all users.</para>
298 <term><option>--no-reload</option></term>
301 <para>When used with <command>enable</command> and
302 <command>disable</command>, do not implicitly reload daemon
303 configuration after executing the changes.</para>
308 <term><option>--no-ask-password</option></term>
311 <para>When used with <command>start</command> and related
312 commands, disables asking for passwords. Background services
313 may require input of a password or passphrase string, for
314 example to unlock system hard disks or cryptographic
315 certificates. Unless this option is specified and the
316 command is invoked from a terminal
317 <command>systemctl</command> will query the user on the
318 terminal for the necessary secrets. Use this option to
319 switch this behavior off. In this case the password must be
320 supplied by some other means (for example graphical password
321 agents) or the service might fail. This also disables
322 querying the user for authentication for privileged
329 <term><option>--kill-who=</option></term>
332 <para>When used with <command>kill</command>, choose which
333 processes to kill. Must be one of <option>main</option>,
334 <option>control</option> or <option>all</option> to select
335 whether to kill only the main process of the unit, the
336 control process or all processes of the unit. If omitted
337 defaults to <option>all</option>.</para>
343 <term><option>-s</option></term>
344 <term><option>--signal=</option></term>
347 <para>When used with <command>kill</command>, choose which
348 signal to send to selected processes. Must be one of the
349 well known signal specifiers such as SIGTERM, SIGINT or
350 SIGSTOP. If omitted defaults to
351 <option>SIGTERM</option>.</para>
356 <term><option>-f</option></term>
357 <term><option>--force</option></term>
360 <para>When used with <command>enable</command>, overwrite
361 any existing conflicting symlinks.</para>
363 <para>When used with <command>halt</command>,
364 <command>poweroff</command>, <command>reboot</command> or
365 <command>kexec</command> execute the selected operation
366 without shutting down all units. However, all processes will
367 be killed forcibly and all file systems are unmounted or
368 remounted read-only. This is hence a drastic but relatively
369 safe option to request an immediate reboot. If
370 <option>--force</option> is specified twice for these
371 operations, they will be executed immediately without
372 terminating any processes or umounting any file
373 systems. Warning: specifying <option>--force</option> twice
374 with any of these operations might result in data
380 <term><option>--root=</option></term>
384 <command>enable</command>/<command>disable</command>/<command>is-enabled</command>
385 (and related commands), use alternative root path when
386 looking for unit files.</para>
392 <term><option>--runtime</option></term>
395 <para>When used with <command>enable</command>,
396 <command>disable</command>, <command>is-enabled</command>
397 (and related commands), make changes only temporarily, so
398 that they are lost on the next reboot. This will have the
399 effect that changes are not made in subdirectories of
400 <filename>/etc</filename> but in <filename>/run</filename>,
401 with identical immediate effects, however, since the latter
402 is lost on reboot, the changes are lost too.</para>
404 <para>Similar, when used with
405 <command>set-cgroup-attr</command>,
406 <command>unset-cgroup-attr</command>,
407 <command>set-cgroup</command> and
408 <command>unset-cgroup</command>, make changes only
409 temporarily, so that they are lost on the next
415 <term><option>-H</option></term>
416 <term><option>--host</option></term>
419 <para>Execute operation remotely. Specify a hostname, or
420 username and hostname separated by @, to connect to. This
421 will use SSH to talk to the remote systemd
427 <term><option>-P</option></term>
428 <term><option>--privileged</option></term>
431 <para>Acquire privileges via PolicyKit before executing the
437 <term><option>-n</option></term>
438 <term><option>--lines=</option></term>
441 <para>When used with <command>status</command> controls the
442 number of journal lines to show, counting from the most
443 recent ones. Takes a positive integer argument. Defaults to
449 <term><option>-o</option></term>
450 <term><option>--output=</option></term>
453 <para>When used with <command>status</command> controls the
454 formatting of the journal entries that are shown. For the
455 available choices see
456 <citerefentry><refentrytitle>journalctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>.
457 Defaults to <literal>short</literal>.</para>
465 <title>Commands</title>
467 <para>The following commands are understood:</para>
471 <term><command>list-units</command></term>
474 <para>List known units (subject to limitations specified
475 with <option>-t</option>).</para>
477 <para>This is the default command.</para>
481 <term><command>start <replaceable>NAME</replaceable>...</command></term>
484 <para>Start (activate) one or more units specified on the
489 <term><command>stop <replaceable>NAME</replaceable>...</command></term>
492 <para>Stop (deactivate) one or more units specified on the
497 <term><command>reload <replaceable>NAME</replaceable>...</command></term>
500 <para>Asks all units listed on the command line to reload
501 their configuration. Note that this will reload the
502 service-specific configuration, not the unit configuration
503 file of systemd. If you want systemd to reload the
504 configuration file of a unit use the
505 <command>daemon-reload</command> command. In other words:
506 for the example case of Apache, this will reload Apache's
507 <filename>httpd.conf</filename> in the web server, not the
508 <filename>apache.service</filename> systemd unit
511 <para>This command should not be confused with the
512 <command>daemon-reload</command> or <command>load</command>
518 <term><command>restart <replaceable>NAME</replaceable>...</command></term>
521 <para>Restart one or more units specified on the command
522 line. If the units are not running yet they will be
527 <term><command>try-restart <replaceable>NAME</replaceable>...</command></term>
530 <para>Restart one or more units specified on the command
531 line if the units are running. Do nothing if units are not
532 running. Note that for compatibility with Red Hat init
533 scripts <command>condrestart</command> is equivalent to this
538 <term><command>reload-or-restart <replaceable>NAME</replaceable>...</command></term>
541 <para>Reload one or more units if they support it. If not,
542 restart them instead. If the units are not running yet they
543 will be started.</para>
547 <term><command>reload-or-try-restart <replaceable>NAME</replaceable>...</command></term>
550 <para>Reload one or more units if they support it. If not,
551 restart them instead. Do nothing if the units are not
552 running. Note that for compatibility with SysV init scripts
553 <command>force-reload</command> is equivalent to this
558 <term><command>isolate <replaceable>NAME</replaceable></command></term>
561 <para>Start the unit specified on the command line and its
562 dependencies and stop all others.</para>
564 <para>This is similar to changing the runlevel in a
565 traditional init system. The <command>isolate</command>
566 command will immediately stop processes that are not enabled
567 in the new unit, possibly including the graphical
568 environment or terminal you are currently using.</para>
570 <para>Note that this is allowed only on units where
571 <option>AllowIsolate=</option> is enabled. See
572 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>
577 <term><command>kill <replaceable>NAME</replaceable>...</command></term>
580 <para>Send a signal to one or more processes of the
581 unit. Use <option>--kill-who=</option> to select which
582 process to kill. Use <option>--kill-mode=</option> to select
583 the kill mode and <option>--signal=</option> to select the
584 signal to send.</para>
588 <term><command>is-active <replaceable>NAME</replaceable>...</command></term>
591 <para>Check whether any of the specified units are active
592 (i.e. running). Returns an exit code 0 if at least one is
593 active, non-zero otherwise. Unless <option>--quiet</option>
594 is specified this will also print the current unit state to
599 <term><command>is-failed <replaceable>NAME</replaceable>...</command></term>
602 <para>Check whether any of the specified units are failed.
603 Returns an exit code 0 if at least one is failed, non-zero
604 otherwise. Unless <option>--quiet</option> is specified this
605 will also print the current unit state to
610 <term><command>status [<replaceable>NAME</replaceable>...|<replaceable>PID</replaceable>...]</command></term>
613 <para>Show terse runtime status information about one or
614 more units, followed by most recent log data from the
615 journal. If no units are specified, show all units (subject
616 to limitations specified with <option>-t</option>). If a PID
617 is passed show information about the unit the process
620 <para>This function is intended to generate human-readable
621 output. If you are looking for computer-parsable output, use
622 <command>show</command> instead.</para>
626 <term><command>show [<replaceable>NAME</replaceable>...|<replaceable>JOB</replaceable>...]</command></term>
629 <para>Show properties of one or more units, jobs, or the
630 manager itself. If no argument is specified properties of
631 the manager will be shown. If a unit name is specified
632 properties of the unit is shown, and if a job id is
633 specified properties of the job is shown. By default, empty
634 properties are suppressed. Use <option>--all</option> to
635 show those too. To select specific properties to show use
636 <option>--property=</option>. This command is intended to be
637 used whenever computer-parsable output is required. Use
638 <command>status</command> if you are looking for formatted
639 human-readable output.</para>
644 <term><command>get-cgroup-attr <replaceable>NAME</replaceable> <replaceable>ATTRIBUTE</replaceable>...</command></term>
647 <para>Retrieve the specified control group attributes of the
648 specified unit. Takes a unit name and one or more attribute
649 names such as <literal>cpu.shares</literal>. This will
650 output the current values of the specified attributes,
651 separated by new-lines. For attributes that take list of
652 items the output will be new-line separated, too. This
653 operation will always try to retrieve the data in question
654 from the kernel first, and if that is not available use the
655 configured values instead. Instead of low-level control
656 group attribute names high-level pretty names may be used,
657 as used for unit execution environment configuration, see
658 <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>
659 for details. For example, passing
660 <literal>memory.limit_in_bytes</literal> and
661 <literal>MemoryLimit</literal> is equivalent.</para>
666 <term><command>set-cgroup-attr <replaceable>NAME</replaceable> <replaceable>ATTRIBUTE</replaceable> <replaceable>VALUE</replaceable>...</command></term>
669 <para>Set the specified control group attribute of the
670 specified unit to the specified value. Takes a unit
671 name and an attribute name such as
672 <literal>cpu.shares</literal>, plus one or more values
673 (multiple values may only be used for attributes that take
674 multiple values). This operation will immediately update the
675 kernel attribute for this unit and persistently store this
676 setting for later reboots (unless <option>--runtime</option>
677 is passed, in which case the setting is not saved
678 persistently and only valid until the next reboot.) Instead
679 of low-level control group attribute names high-level pretty
680 names may be used, as used for unit execution environment
682 <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>
683 for details. For example, passing
684 <literal>memory.limit_in_bytes</literal> and
685 <literal>MemoryLimit</literal> is equivalent. This operation
686 will implicitly create a control group for the unit in the
687 controller the attribute belongs to, if needed. For
688 attributes that take multiple values, this operation will
689 append the specified values to the previously set values
690 list (use <command>unset-cgroup-attr</command> to reset the
691 list explicitly). For attributes that take a single value
692 only the list will be reset implicitly.</para>
697 <term><command>unset-cgroup-attr <replaceable>NAME</replaceable> <replaceable>ATTRIBUTE</replaceable>...</command></term>
699 <listitem><para>Unset the specified control group attributes
700 of the specified unit. Takes a unit name and one or more
701 attribut names such as <literal>cpu.shares</literal>. This
702 operation might or might not have an immediate effect on the
703 current kernel attribute value. This will remove any
704 persistently stored configuration values for this attribute
705 (as set with <command>set-cgroup-attr</command> before),
706 unless <option>--runtime</option> is passed, in which case the
707 configuration is reset only until the next reboot. Again,
708 high-level control group attributes may be used instead of the
709 low-level kernel ones. For attributes which take multiple
710 values, all currently set values are reset.</para>
715 <term><command>set-cgroup <replaceable>NAME</replaceable> <replaceable>CGROUP</replaceable>...</command></term>
716 <term><command>unset-cgroup <replaceable>NAME</replaceable> <replaceable>CGROUP</replaceable>...</command></term>
718 <listitem><para>Add or remove a unit to/from a specific
719 control group hierarchy and/or control group path. Takes a
720 unit name, plus a control group specification in the syntax
721 <replaceable>CONTROLLER</replaceable>:<replaceable>PATH</replaceable>
722 or <replaceable>CONTROLLER</replaceable>. In the latter syntax
723 (where the path is ommitted) the default unit control group
724 path is implied. Examples: <literal>cpu</literal> or
725 <literal>cpu:/foo/bar</literal>. If a unit is removed from a
726 control group hierarchy all its processes will be moved to the
727 root group of the hierarchy and all control group attributes
728 will be reset. These operations are immediately reflected in
729 the kernel hierarchy, and stored persistently to disk (unless
730 <option>--runtime</option> is passed).</para>
735 <term><command>help <replaceable>NAME</replaceable>...|<replaceable>PID</replaceable>...</command></term>
738 <para>Show manual pages for one or more units, if
739 available. If a PID is passed the manual pages for the unit
740 the process of the PID belongs to is
746 <term><command>reset-failed [<replaceable>NAME</replaceable>...]</command></term>
749 <para>Reset the <literal>failed</literal> state of the
750 specified units, or if no unit name is passed of all
751 units. When a unit fails in some way (i.e. process exiting
752 with non-zero error code, terminating abnormally or timing
753 out) it will automatically enter the
754 <literal>failed</literal> state and its exit code and status
755 is recorded for introspection by the administrator until the
756 service is restarted or reset with this command.</para>
761 <term><command>list-unit-files</command></term>
764 <para>List installed unit files.</para>
769 <term><command>enable <replaceable>NAME</replaceable>...</command></term>
772 <para>Enable one or more unit files or unit file instances,
773 as specified on the command line. This will create a number
774 of symlinks as encoded in the <literal>[Install]</literal>
775 sections of the unit files. After the symlinks have been
776 created the systemd configuration is reloaded (in a way that
777 is equivalent to <command>daemon-reload</command>) to ensure
778 the changes are taken into account immediately. Note that
779 this does not have the effect that any of the units enabled
780 are also started at the same time. If this is desired a
781 separate <command>start</command> command must be invoked
782 for the unit. Also note that in case of instance enablement,
783 symlinks named same as instances are created in install
784 location, however they all point to the same template unit
787 <para>This command will print the actions executed. This
788 output may be suppressed by passing <option>--quiet</option>.
791 <para>Note that this operation creates only the suggested
792 symlinks for the units. While this command is the
793 recommended way to manipulate the unit configuration
794 directory, the administrator is free to make additional
795 changes manually, by placing or removing symlinks in the
796 directory. This is particularly useful to create
797 configurations that deviate from the suggested default
798 installation. In this case the administrator must make sure
799 to invoke <command>daemon-reload</command> manually as
800 necessary, to ensure his changes are taken into account.
803 <para>Enabling units should not be confused with starting
804 (activating) units, as done by the <command>start</command>
805 command. Enabling and starting units is orthogonal: units
806 may be enabled without being started and started without
807 being enabled. Enabling simply hooks the unit into various
808 suggested places (for example, so that the unit is
809 automatically started on boot or when a particular kind of
810 hardware is plugged in). Starting actually spawns the daemon
811 process (in case of service units), or binds the socket (in
812 case of socket units), and so on.</para>
814 <para>Depending on whether <option>--system</option>,
815 <option>--user</option> or <option>--global</option> is
816 specified this enables the unit for the system, for the
817 calling user only or for all future logins of all
818 users. Note that in the last case no systemd daemon
819 configuration is reloaded.</para>
824 <term><command>disable <replaceable>NAME</replaceable>...</command></term>
827 <para>Disables one or more units. This removes all symlinks
828 to the specified unit files from the unit configuration
829 directory, and hence undoes the changes made by
830 <command>enable</command>. Note however that this removes
831 all symlinks to the unit files (i.e. including manual
832 additions), not just those actually created by
833 <command>enable</command>. This call implicitly reloads the
834 systemd daemon configuration after completing the disabling
835 of the units. Note that this command does not implicitly
836 stop the units that are being disabled. If this is desired
837 an additional <command>stop</command> command should be
838 executed afterwards.</para>
840 <para>This command will print the actions executed. This
841 output may be suppressed by passing <option>--quiet</option>.
844 <para>This command honors <option>--system</option>,
845 <option>--user</option>, <option>--global</option> in a
846 similar way as <command>enable</command>.</para>
851 <term><command>is-enabled <replaceable>NAME</replaceable>...</command></term>
854 <para>Checks whether any of the specified unit files are
855 enabled (as with <command>enable</command>). Returns an exit
856 code of 0 if at least one is enabled, non-zero
857 otherwise. Prints the current enable status. To suppress
858 this output use <option>--quiet</option>.</para>
863 <term><command>reenable <replaceable>NAME</replaceable>...</command></term>
866 <para>Reenable one or more unit files, as specified on the
867 command line. This is a combination of
868 <command>disable</command> and <command>enable</command> and
869 is useful to reset the symlinks a unit is enabled with to
870 the defaults configured in the <literal>[Install]</literal>
871 section of the unit file.</para>
876 <term><command>preset <replaceable>NAME</replaceable>...</command></term>
879 <para>Reset one or more unit files, as specified on the
880 command line, to the defaults configured in the preset
881 policy files. This has the same effect as
882 <command>disable</command> or <command>enable</command>,
883 depending how the unit is listed in the preset files. For
884 more information on preset policy format see
885 <citerefentry><refentrytitle>systemd.preset</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
886 For more information on the concept of presets please
888 <ulink url="http://freedesktop.org/wiki/Software/systemd/Preset">Preset</ulink>
894 <term><command>mask <replaceable>NAME</replaceable>...</command></term>
897 <para>Mask one or more unit files, as specified on the
898 command line. This will link these units to
899 <filename>/dev/null</filename>, making it impossible to
900 start them. This is a stronger version of
901 <command>disable</command>, since it prohibits all kinds of
902 activation of the unit, including manual activation. Use
903 this option with care.</para>
908 <term><command>unmask <replaceable>NAME</replaceable>...</command></term>
911 <para>Unmask one or more unit files, as specified on the
912 command line. This will undo the effect of
913 <command>mask</command>.</para>
918 <term><command>link <replaceable>FILENAME</replaceable>...</command></term>
921 <para>Link a unit file that is not in the unit file search
922 paths into the unit file search path. This requires an
923 absolute path to a unit file. The effect of this can be
924 undone with <command>disable</command>. The effect of this
925 command is that a unit file is available for
926 <command>start</command> and other commands although it
927 isn't installed directly in the unit search path.</para>
932 <term><command>load <replaceable>NAME</replaceable>...</command></term>
935 <para>Load one or more units specified on the command
936 line. This will simply load their configuration from disk,
937 but not start them. To start them you need to use the
938 <command>start</command> command which will implicitly load
939 a unit that has not been loaded yet. Note that systemd
940 garbage collects loaded units that are not active or
941 referenced by an active unit. This means that units loaded
942 this way will usually not stay loaded for long. Also note
943 that this command cannot be used to reload unit
944 configuration. Use the <command>daemon-reload</command>
945 command for that. All in all, this command is of little use
946 except for debugging.</para>
948 <para>This command should not be confused with the
949 <command>daemon-reload</command> or
950 <command>reload</command>.</para>
954 <term><command>list-jobs</command></term>
957 <para>List jobs that are in progress.</para>
961 <term><command>cancel <replaceable>JOB</replaceable>...</command></term>
964 <para>Cancel one or more jobs specified on the command line
965 by their numeric job IDs. If no job id is specified, cancel
966 all pending jobs.</para>
970 <term><command>dump</command></term>
973 <para>Dump server status. This will output a (usually very
974 long) human readable manager status dump. Its format is
975 subject to change without notice and should not be parsed by
980 <term><command>list-dependencies <replaceable>NAME</replaceable></command></term>
983 <para>Shows required and wanted units of the specified
984 unit. If no unit is specified
985 <filename>default.target</filename> is implied. Target units
986 are recursively expanded. When <option>--all</option> is
987 passed all other units are recursively expanded as
992 <term><command>snapshot [<replaceable>NAME</replaceable>]</command></term>
995 <para>Create a snapshot. If a snapshot name is specified,
996 the new snapshot will be named after it. If none is
997 specified an automatic snapshot name is generated. In either
998 case, the snapshot name used is printed to STDOUT, unless
999 <option>--quiet</option> is specified.</para>
1001 <para>A snapshot refers to a saved state of the systemd
1002 manager. It is implemented itself as a unit that is
1003 generated dynamically with this command and has dependencies
1004 on all units active at the time. At a later time the user
1005 may return to this state by using the
1006 <command>isolate</command> command on the snapshot unit.
1009 <para>Snapshots are only useful for saving and restoring
1010 which units are running or are stopped, they do not
1011 save/restore any other state. Snapshots are dynamic and lost
1016 <term><command>delete <replaceable>NAME</replaceable>...</command></term>
1019 <para>Remove a snapshot previously created with
1020 <command>snapshot</command>.</para>
1024 <term><command>daemon-reload</command></term>
1027 <para>Reload systemd manager configuration. This will reload
1028 all unit files and recreate the entire dependency
1029 tree. While the daemon is reloaded, all sockets systemd
1030 listens on on behalf of user configuration will stay
1031 accessible.</para> <para>This command should not be confused
1032 with the <command>load</command> or
1033 <command>reload</command> commands.</para>
1037 <term><command>daemon-reexec</command></term>
1040 <para>Reexecute the systemd manager. This will serialize the
1041 manager state, reexecute the process and deserialize the
1042 state again. This command is of little use except for
1043 debugging and package upgrades. Sometimes it might be
1044 helpful as a heavy-weight <command>daemon-reload</command>.
1045 While the daemon is reexecuted all sockets systemd listens
1046 on on behalf of user configuration will stay accessible.
1051 <term><command>show-environment</command></term>
1054 <para>Dump the systemd manager environment block. The
1055 environment block will be dumped in straight-forward form
1056 suitable for sourcing into a shell script. This environment
1057 block will be passed to all processes the manager
1062 <term><command>set-environment <replaceable>VARIABLE=VALUE</replaceable>...</command></term>
1065 <para>Set one or more systemd manager environment variables,
1066 as specified on the command line.</para>
1070 <term><command>unset-environment <replaceable>VARIABLE</replaceable>...</command></term>
1073 <para>Unset one or more systemd manager environment
1074 variables. If only a variable name is specified it will be
1075 removed regardless of its value. If a variable and a value
1076 are specified the variable is only removed if it has the
1077 specified value.</para>
1081 <term><command>default</command></term>
1084 <para>Enter default mode. This is mostly equivalent to
1085 <command>isolate default.target</command>.</para>
1089 <term><command>rescue</command></term>
1092 <para>Enter rescue mode. This is mostly equivalent to
1093 <command>isolate rescue.target</command> but also prints a
1094 wall message to all users.</para>
1098 <term><command>emergency</command></term>
1101 <para>Enter emergency mode. This is mostly equivalent to
1102 <command>isolate emergency.target</command> but also prints
1103 a wall message to all users.</para>
1107 <term><command>halt</command></term>
1110 <para>Shut down and halt the system. This is mostly equivalent to
1111 <command>start halt.target --irreversible</command> but also
1112 prints a wall message to all users. If combined with
1113 <option>--force</option> shutdown of all running services is
1114 skipped, however all processes are killed and all file
1115 systems are unmounted or mounted read-only, immediately
1116 followed by the system halt. If <option>--force</option> is
1117 specified twice the operation is immediately executed
1118 without terminating any processes or unmounting any file
1119 systems. This may result in data loss.</para>
1123 <term><command>poweroff</command></term>
1126 <para>Shut down and power-off the system. This is mostly
1127 equivalent to <command>start poweroff.target --irreversible</command>
1128 but also prints a wall message to all users. If combined with
1129 <option>--force</option> shutdown of all running services is
1130 skipped, however all processes are killed and all file
1131 systems are unmounted or mounted read-only, immediately
1132 followed by the powering off. If <option>--force</option> is
1133 specified twice the operation is immediately executed
1134 without terminating any processes or unmounting any file
1135 systems. This may result in data loss.</para>
1139 <term><command>reboot</command></term>
1142 <para>Shut down and reboot the system. This is mostly
1143 equivalent to <command>start reboot.target --irreversible</command>
1144 but also prints a wall message to all users. If combined with
1145 <option>--force</option> shutdown of all running services is
1146 skipped, however all processes are killed and all file
1147 systems are unmounted or mounted read-only, immediately
1148 followed by the reboot. If <option>--force</option> is
1149 specified twice the operation is immediately executed
1150 without terminating any processes or unmounting any file
1151 systems. This may result in data loss.</para>
1155 <term><command>kexec</command></term>
1158 <para>Shut down and reboot the system via kexec. This is
1159 mostly equivalent to <command>start kexec.target --irreversible</command>
1160 but also prints a wall message to all users. If combined
1161 with <option>--force</option> shutdown of all running
1162 services is skipped, however all processes are killed and
1163 all file systems are unmounted or mounted read-only,
1164 immediately followed by the reboot.</para>
1168 <term><command>exit</command></term>
1171 <para>Ask the systemd manager to quit. This is only
1172 supported for user service managers (i.e. in conjunction
1173 with the <option>--user</option> option) and will fail
1179 <term><command>suspend</command></term>
1182 <para>Suspend the system. This will trigger activation of
1183 the special <filename>suspend.target</filename> target.
1188 <term><command>hibernate</command></term>
1191 <para>Hibernate the system. This will trigger activation of
1192 the special <filename>hibernate.target</filename> target.
1197 <term><command>hybrid-sleep</command></term>
1200 <para>Hibernate and suspend the system. This will trigger
1201 activation of the special
1202 <filename>hybrid-sleep.target</filename> target.</para>
1206 <term><command>switch-root <replaceable>ROOT</replaceable> [<replaceable>INIT</replaceable>]</command></term>
1209 <para>Switches to a different root directory and executes a
1210 new system manager process below it. This is intended for
1211 usage in initial RAM disks ("initrd"), and will transition
1212 from the initrd's system manager process (a.k.a "init"
1213 process) to the main system manager process. Takes two
1214 arguments: the directory to make the new root directory, and
1215 the path to the new system manager binary below it to
1216 execute as PID 1. If the latter is omitted or the empty
1217 string, a systemd binary will automatically be searched for
1218 and used as init. If the system manager path is omitted or
1219 equal to the empty string the state of the initrd's system
1220 manager process is passed to the main system manager, which
1221 allows later introspection of the state of the services
1222 involved in the initrd boot.</para>
1230 <title>Exit status</title>
1232 <para>On success 0 is returned, a non-zero failure
1233 code otherwise.</para>
1237 <title>Environment</title>
1239 <variablelist class='environment-variables'>
1241 <term><varname>$SYSTEMD_PAGER</varname></term>
1244 <para>Pager to use when <option>--no-pager</option> is not
1245 given; overrides <varname>$PAGER</varname>. Setting this to
1246 an empty string or the value <literal>cat</literal> is
1247 equivalent to passing
1248 <option>--no-pager</option>.</para>
1255 <title>See Also</title>
1257 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1258 <citerefentry><refentrytitle>systemadm</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1259 <citerefentry><refentrytitle>journalctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1260 <citerefentry><refentrytitle>loginctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1261 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1262 <citerefentry><refentrytitle>systemd.special</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
1263 <citerefentry><refentrytitle>wall</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1264 <citerefentry><refentrytitle>systemd.preset</refentrytitle><manvolnum>5</manvolnum></citerefentry>