1SYSTEMD.SERVICE(5) systemd.service SYSTEMD.SERVICE(5)
2
6 systemd.service - Service unit configuration
7
9 service.service
10
12 A unit configuration file whose name ends in ".service" encodes
13 information about a process controlled and supervised by systemd.
14 This man page lists the configuration options specific to this unit
16 type. See systemd.unit(5) for the common options of all unit
17 configuration files. The common configuration items are configured in
18 the generic "[Unit]" and "[Install]" sections. The service specific
19 configuration options are configured in the "[Service]" section.
20 Additional options are listed in systemd.exec(5), which define the
22 execution environment the commands are executed in, and in
23 systemd.kill(5), which define the way the processes of the service are
24 terminated, and in systemd.resource-control(5), which configure
25 resource control settings for the processes of the service.
26 If a service is requested under a certain name but no unit
28 configuration file is found, systemd looks for a SysV init script by
29 the same name (with the .service suffix removed) and dynamically
30 creates a service unit from that script. This is useful for
31 compatibility with SysV. Note that this compatibility is quite
32 comprehensive but not 100%. For details about the incompatibilities,
33 see the Incompatibilities with SysV[1] document.
34 The systemd-run(1) command allows creating .service and .scope units
36 dynamically and transiently from the command line.
37
39 It is possible for systemd services to take a single argument via the
40 "service@argument.service" syntax. Such services are called
41 "instantiated" services, while the unit definition without the argument
42 parameter is called a "template". An example could be a dhcpcd@.service
43 service template which takes a network interface as a parameter to form
44 an instantiated service. Within the service file, this parameter or
45 "instance name" can be accessed with %-specifiers. See systemd.unit(5)
46 for details.
47
49 Implicit Dependencies
50 The following dependencies are implicitly added:
51 · Services with Type=dbus set automatically acquire dependencies of
53 type Requires= and After= on dbus.socket.
54 · Socket activated services are automatically ordered after their
56 activating .socket units via an automatic After= dependency.
57 Services also pull in all .socket units listed in Sockets= via
58 automatic Wants= and After= dependencies.
59 Additional implicit dependencies may be added as result of execution
61 and resource control parameters as documented in systemd.exec(5) and
62 systemd.resource-control(5).
63 Default Dependencies
65 The following dependencies are added unless DefaultDependencies=no is
66 set:
67 · Service units will have dependencies of type Requires= and After=
69 on sysinit.target, a dependency of type After= on basic.target as
70 well as dependencies of type Conflicts= and Before= on
71 shutdown.target. These ensure that normal service units pull in
72 basic system initialization, and are terminated cleanly prior to
73 system shutdown. Only services involved with early boot or late
74 system shutdown should disable this option.
75 · Instanced service units (i.e. service units with an "@" in their
77 name) are assigned by default a per-template slice unit (see
78 systemd.slice(5)), named after the template unit, containing all
79 instances of the specific template. This slice is normally stopped
80 at shutdown, together with all template instances. If that is not
81 desired, set DefaultDependencies=no in the template unit, and
82 either define your own per-template slice unit file that also sets
83 DefaultDependencies=no, or set Slice=system.slice (or another
84 suitable slice) in the template unit. Also see systemd.resource-
85 control(5).
86
88 Service files must include a "[Service]" section, which carries
89 information about the service and the process it supervises. A number
90 of options that may be used in this section are shared with other unit
91 types. These options are documented in systemd.exec(5), systemd.kill(5)
92 and systemd.resource-control(5). The options specific to the
93 "[Service]" section of service units are the following:
94 Type=
96 Configures the process start-up type for this service unit. One of
97 simple, exec, forking, oneshot, dbus, notify or idle:
98 · If set to simple (the default if ExecStart= is specified but
100 neither Type= nor BusName= are), the service manager will
101 consider the unit started immediately after the main service
102 process has been forked off. It is expected that the process
103 configured with ExecStart= is the main process of the service.
104 In this mode, if the process offers functionality to other
105 processes on the system, its communication channels should be
106 installed before the service is started up (e.g. sockets set up
107 by systemd, via socket activation), as the service manager will
108 immediately proceed starting follow-up units, right after
109 creating the main service process, and before executing the
110 service's binary. Note that this means systemctl start command
111 lines for simple services will report success even if the
112 service's binary cannot be invoked successfully (for example
113 because the selected User= doesn't exist, or the service binary
114 is missing).
115 · The exec type is similar to simple, but the service manager
117 will consider the unit started immediately after the main
118 service binary has been executed. The service manager will
119 delay starting of follow-up units until that point. (Or in
120 other words: simple proceeds with further jobs right after
121 fork() returns, while exec will not proceed before both fork()
122 and execve() in the service process succeeded.) Note that this
123 means systemctl start command lines for exec services will
124 report failure when the service's binary cannot be invoked
125 successfully (for example because the selected User= doesn't
126 exist, or the service binary is missing).
127 · If set to forking, it is expected that the process configured
129 with ExecStart= will call fork() as part of its start-up. The
130 parent process is expected to exit when start-up is complete
131 and all communication channels are set up. The child continues
132 to run as the main service process, and the service manager
133 will consider the unit started when the parent process exits.
134 This is the behavior of traditional UNIX services. If this
135 setting is used, it is recommended to also use the PIDFile=
136 option, so that systemd can reliably identify the main process
137 of the service. systemd will proceed with starting follow-up
138 units as soon as the parent process exits.
139 · Behavior of oneshot is similar to simple; however, the service
141 manager will consider the unit up after the main process exits.
142 It will then start follow-up units. RemainAfterExit= is
143 particularly useful for this type of service. Type=oneshot is
144 the implied default if neither Type= nor ExecStart= are
145 specified. Note that if this option is used without
146 RemainAfterExit= the service will never enter "active" unit
147 state, but directly transition from "activating" to
148 "deactivating" or "dead" since no process is configured that
149 shall run continously. In particular this means that after a
150 service of this type ran (and which has RemainAfterExit= not
151 set) it will not show up as started afterwards, but as dead.
152 · Behavior of dbus is similar to simple; however, it is expected
154 that the service acquires a name on the D-Bus bus, as
155 configured by BusName=. systemd will proceed with starting
156 follow-up units after the D-Bus bus name has been acquired.
157 Service units with this option configured implicitly gain
158 dependencies on the dbus.socket unit. This type is the default
159 if BusName= is specified.
160 · Behavior of notify is similar to exec; however, it is expected
162 that the service sends a notification message via sd_notify(3)
163 or an equivalent call when it has finished starting up. systemd
164 will proceed with starting follow-up units after this
165 notification message has been sent. If this option is used,
166 NotifyAccess= (see below) should be set to open access to the
167 notification socket provided by systemd. If NotifyAccess= is
168 missing or set to none, it will be forcibly set to main. Note
169 that currently Type=notify will not work if used in combination
170 with PrivateNetwork=yes.
171 · Behavior of idle is very similar to simple; however, actual
173 execution of the service program is delayed until all active
174 jobs are dispatched. This may be used to avoid interleaving of
175 output of shell services with the status output on the console.
176 Note that this type is useful only to improve console output,
177 it is not useful as a general unit ordering tool, and the
178 effect of this service type is subject to a 5s timeout, after
179 which the service program is invoked anyway.
180 It is generally recommended to use Type=simple for long-running
182 services whenever possible, as it is the simplest and fastest
183 option. However, as this service type won't propagate service
184 start-up failures and doesn't allow ordering of other units against
185 completion of initialization of the service (which for example is
186 useful if clients need to connect to the service through some form
187 of IPC, and the IPC channel is only established by the service
188 itself — in contrast to doing this ahead of time through socket or
189 bus activation or similar), it might not be sufficient for many
190 cases. If so, notify or dbus (the latter only in case the service
191 provides a D-Bus interface) are the preferred options as they allow
192 service program code to precisely schedule when to consider the
193 service started up successfully and when to proceed with follow-up
194 units. The notify service type requires explicit support in the
195 service codebase (as sd_notify() or an equivalent API needs to be
196 invoked by the service at the appropriate time) — if it's not
197 supported, then forking is an alternative: it supports the
198 traditional UNIX service start-up protocol. Finally, exec might be
199 an option for cases where it is enough to ensure the service binary
200 is invoked, and where the service binary itself executes no or
201 little initialization on its own (and its initialization is
202 unlikely to fail). Note that using any type other than simple
203 possibly delays the boot process, as the service manager needs to
204 wait for service initialization to complete. It is hence
205 recommended not to needlessly use any types other than simple.
206 (Also note it is generally not recommended to use idle or oneshot
207 for long-running services.)
208 RemainAfterExit=
210 Takes a boolean value that specifies whether the service shall be
211 considered active even when all its processes exited. Defaults to
212 no.
213 GuessMainPID=
215 Takes a boolean value that specifies whether systemd should try to
216 guess the main PID of a service if it cannot be determined
217 reliably. This option is ignored unless Type=forking is set and
218 PIDFile= is unset because for the other types or with an explicitly
219 configured PID file, the main PID is always known. The guessing
220 algorithm might come to incorrect conclusions if a daemon consists
221 of more than one process. If the main PID cannot be determined,
222 failure detection and automatic restarting of a service will not
223 work reliably. Defaults to yes.
224 PIDFile=
226 Takes a path referring to the PID file of the service. Usage of
227 this option is recommended for services where Type= is set to
228 forking. The path specified typically points to a file below /run/.
229 If a relative path is specified it is hence prefixed with /run/.
230 The service manager will read the PID of the main process of the
231 service from this file after start-up of the service. The service
232 manager will not write to the file configured here, although it
233 will remove the file after the service has shut down if it still
234 exists. The PID file does not need to be owned by a privileged
235 user, but if it is owned by an unprivileged user additional safety
236 restrictions are enforced: the file may not be a symlink to a file
237 owned by a different user (neither directly nor indirectly), and
238 the PID file must refer to a process already belonging to the
239 service.
240 BusName=
242 Takes a D-Bus bus name that this service is reachable as. This
243 option is mandatory for services where Type= is set to dbus.
244 ExecStart=
246 Commands with their arguments that are executed when this service
247 is started. The value is split into zero or more command lines
248 according to the rules described below (see section "Command Lines"
249 below).
250 Unless Type= is oneshot, exactly one command must be given. When
252 Type=oneshot is used, zero or more commands may be specified.
253 Commands may be specified by providing multiple command lines in
254 the same directive, or alternatively, this directive may be
255 specified more than once with the same effect. If the empty string
256 is assigned to this option, the list of commands to start is reset,
257 prior assignments of this option will have no effect. If no
258 ExecStart= is specified, then the service must have
259 RemainAfterExit=yes and at least one ExecStop= line set. (Services
260 lacking both ExecStart= and ExecStop= are not valid.)
261 For each of the specified commands, the first argument must be
263 either an absolute path to an executable or a simple file name
264 without any slashes. Optionally, this filename may be prefixed with
265 a number of special characters:
266 Table 1. Special executable prefixes
268 ┌───────┬────────────────────────────┐
269 │Prefix │ Effect │
270 ├───────┼────────────────────────────┤
271 │"@" │ If the executable path is │
272 │ │ prefixed with "@", the │
273 │ │ second specified token │
274 │ │ will be passed as │
275 │ │ "argv[0]" to the executed │
276 │ │ process (instead of the │
277 │ │ actual filename), followed │
278 │ │ by the further arguments │
279 │ │ specified. │
280 ├───────┼────────────────────────────┤
281 │"-" │ If the executable path is │
282 │ │ prefixed with "-", an exit │
283 │ │ code of the command │
284 │ │ normally considered a │
285 │ │ failure (i.e. non-zero │
286 │ │ exit status or abnormal │
287 │ │ exit due to signal) is │
288 │ │ recorded, but has no │
289 │ │ further effect and is │
290 │ │ considered equivalent to │
291 │ │ success. │
292 ├───────┼────────────────────────────┤
293 │":" │ If the executable path is │
294 │ │ prefixed with ":", │
295 │ │ environment variable │
296 │ │ substitution (as described │
297 │ │ by the "Command Lines" │
298 │ │ section below) is not │
299 │ │ applied. │
300 ├───────┼────────────────────────────┤
301 │"+" │ If the executable path is │
302 │ │ prefixed with "+" then the │
303 │ │ process is executed with │
304 │ │ full privileges. In this │
305 │ │ mode privilege │
306 │ │ restrictions configured │
307 │ │ with User=, Group=, │
308 │ │ CapabilityBoundingSet= or │
309 │ │ the various file system │
310 │ │ namespacing options (such │
311 │ │ as PrivateDevices=, │
312 │ │ PrivateTmp=) are not │
313 │ │ applied to the invoked │
314 │ │ command line (but still │
315 │ │ affect any other │
316 │ │ ExecStart=, ExecStop=, ... │
317 │ │ lines). │
318 ├───────┼────────────────────────────┤
319 │"!" │ Similar to the "+" │
320 │ │ character discussed above │
321 │ │ this permits invoking │
322 │ │ command lines with │
323 │ │ elevated privileges. │
324 │ │ However, unlike "+" the │
325 │ │ "!" character exclusively │
326 │ │ alters the effect of │
327 │ │ User=, Group= and │
328 │ │ SupplementaryGroups=, i.e. │
329 │ │ only the stanzas that │
330 │ │ affect user and group │
331 │ │ credentials. Note that │
332 │ │ this setting may be │
333 │ │ combined with │
334 │ │ DynamicUser=, in which │
335 │ │ case a dynamic user/group │
336 │ │ pair is allocated before │
337 │ │ the command is invoked, │
338 │ │ but credential changing is │
339 │ │ left to the executed │
340 │ │ process itself. │
341 ├───────┼────────────────────────────┤
342 │"!!" │ This prefix is very │
343 │ │ similar to "!", however it │
344 │ │ only has an effect on │
345 │ │ systems lacking support │
346 │ │ for ambient process │
347 │ │ capabilities, i.e. without │
348 │ │ support for │
349 │ │ AmbientCapabilities=. It's │
350 │ │ intended to be used for │
351 │ │ unit files that take │
352 │ │ benefit of ambient │
353 │ │ capabilities to run │
354 │ │ processes with minimal │
355 │ │ privileges wherever │
356 │ │ possible while remaining │
357 │ │ compatible with systems │
358 │ │ that lack ambient │
359 │ │ capabilities support. Note │
360 │ │ that when "!!" is used, │
361 │ │ and a system lacking │
362 │ │ ambient capability support │
363 │ │ is detected any configured │
364 │ │ SystemCallFilter= and │
365 │ │ CapabilityBoundingSet= │
366 │ │ stanzas are implicitly │
367 │ │ modified, in order to │
368 │ │ permit spawned processes │
369 │ │ to drop credentials and │
370 │ │ capabilities themselves, │
371 │ │ even if this is configured │
372 │ │ to not be allowed. │
373 │ │ Moreover, if this prefix │
374 │ │ is used and a system │
375 │ │ lacking ambient capability │
376 │ │ support is detected │
377 │ │ AmbientCapabilities= will │
378 │ │ be skipped and not be │
379 │ │ applied. On systems │
380 │ │ supporting ambient │
381 │ │ capabilities, "!!" has no │
382 │ │ effect and is redundant. │
383 └───────┴────────────────────────────┘
384 "@", "-", ":", and one of "+"/"!"/"!!" may be used together and
385 they can appear in any order. However, only one of "+", "!", "!!"
386 may be used at a time. Note that these prefixes are also supported
387 for the other command line settings, i.e. ExecStartPre=,
388 ExecStartPost=, ExecReload=, ExecStop= and ExecStopPost=.
389 If more than one command is specified, the commands are invoked
391 sequentially in the order they appear in the unit file. If one of
392 the commands fails (and is not prefixed with "-"), other lines are
393 not executed, and the unit is considered failed.
394 Unless Type=forking is set, the process started via this command
396 line will be considered the main process of the daemon.
397 ExecStartPre=, ExecStartPost=
399 Additional commands that are executed before or after the command
400 in ExecStart=, respectively. Syntax is the same as for ExecStart=,
401 except that multiple command lines are allowed and the commands are
402 executed one after the other, serially.
403 If any of those commands (not prefixed with "-") fail, the rest are
405 not executed and the unit is considered failed.
406 ExecStart= commands are only run after all ExecStartPre= commands
408 that were not prefixed with a "-" exit successfully.
409 ExecStartPost= commands are only run after the commands specified
411 in ExecStart= have been invoked successfully, as determined by
412 Type= (i.e. the process has been started for Type=simple or
413 Type=idle, the last ExecStart= process exited successfully for
414 Type=oneshot, the initial process exited successfully for
415 Type=forking, "READY=1" is sent for Type=notify, or the BusName=
416 has been taken for Type=dbus).
417 Note that ExecStartPre= may not be used to start long-running
419 processes. All processes forked off by processes invoked via
420 ExecStartPre= will be killed before the next service process is
421 run.
422 Note that if any of the commands specified in ExecStartPre=,
424 ExecStart=, or ExecStartPost= fail (and are not prefixed with "-",
425 see above) or time out before the service is fully up, execution
426 continues with commands specified in ExecStopPost=, the commands in
427 ExecStop= are skipped.
428 ExecCondition=
430 Optional commands that are executed before the command(s) in
431 ExecStartPre=. Syntax is the same as for ExecStart=, except that
432 multiple command lines are allowed and the commands are executed
433 one after the other, serially.
434 The behavior is like an ExecStartPre= and condition check hybrid:
436 when an ExecCondition= command exits with exit code 1 through 254
437 (inclusive), the remaining commands are skipped and the unit is not
438 marked as failed. However, if an ExecCondition= command exits with
439 255 or abnormally (e.g. timeout, killed by a signal, etc.), the
440 unit will be considered failed (and remaining commands will be
441 skipped). Exit code of 0 or those matching SuccessExitStatus= will
442 continue execution to the next command(s).
443 The same recommendations about not running long-running processes
445 in ExecStartPre= also applies to ExecCondition=. ExecCondition=
446 will also run the commands in ExecStopPost=, as part of stopping
447 the service, in the case of any non-zero or abnormal exits, like
448 the ones described above.
449 ExecReload=
451 Commands to execute to trigger a configuration reload in the
452 service. This argument takes multiple command lines, following the
453 same scheme as described for ExecStart= above. Use of this setting
454 is optional. Specifier and environment variable substitution is
455 supported here following the same scheme as for ExecStart=.
456 One additional, special environment variable is set: if known,
458 $MAINPID is set to the main process of the daemon, and may be used
459 for command lines like the following:
460 /bin/kill -HUP $MAINPID
462 Note however that reloading a daemon by sending a signal (as with
464 the example line above) is usually not a good choice, because this
465 is an asynchronous operation and hence not suitable to order
466 reloads of multiple services against each other. It is strongly
467 recommended to set ExecReload= to a command that not only triggers
468 a configuration reload of the daemon, but also synchronously waits
469 for it to complete.
470 ExecStop=
472 Commands to execute to stop the service started via ExecStart=.
473 This argument takes multiple command lines, following the same
474 scheme as described for ExecStart= above. Use of this setting is
475 optional. After the commands configured in this option are run, it
476 is implied that the service is stopped, and any processes remaining
477 for it are terminated according to the KillMode= setting (see
478 systemd.kill(5)). If this option is not specified, the process is
479 terminated by sending the signal specified in KillSignal= when
480 service stop is requested. Specifier and environment variable
481 substitution is supported (including $MAINPID, see above).
482 Note that it is usually not sufficient to specify a command for
484 this setting that only asks the service to terminate (for example,
485 by queuing some form of termination signal for it), but does not
486 wait for it to do so. Since the remaining processes of the services
487 are killed according to KillMode= and KillSignal= as described
488 above immediately after the command exited, this may not result in
489 a clean stop. The specified command should hence be a synchronous
490 operation, not an asynchronous one.
491 Note that the commands specified in ExecStop= are only executed
493 when the service started successfully first. They are not invoked
494 if the service was never started at all, or in case its start-up
495 failed, for example because any of the commands specified in
496 ExecStart=, ExecStartPre= or ExecStartPost= failed (and weren't
497 prefixed with "-", see above) or timed out. Use ExecStopPost= to
498 invoke commands when a service failed to start up correctly and is
499 shut down again. Also note that the stop operation is always
500 performed if the service started successfully, even if the
501 processes in the service terminated on their own or were killed.
502 The stop commands must be prepared to deal with that case.
503 $MAINPID will be unset if systemd knows that the main process
504 exited by the time the stop commands are called.
505 Service restart requests are implemented as stop operations
507 followed by start operations. This means that ExecStop= and
508 ExecStopPost= are executed during a service restart operation.
509 It is recommended to use this setting for commands that communicate
511 with the service requesting clean termination. For post-mortem
512 clean-up steps use ExecStopPost= instead.
513 ExecStopPost=
515 Additional commands that are executed after the service is stopped.
516 This includes cases where the commands configured in ExecStop= were
517 used, where the service does not have any ExecStop= defined, or
518 where the service exited unexpectedly. This argument takes multiple
519 command lines, following the same scheme as described for
520 ExecStart=. Use of these settings is optional. Specifier and
521 environment variable substitution is supported. Note that – unlike
522 ExecStop= – commands specified with this setting are invoked when a
523 service failed to start up correctly and is shut down again.
524 It is recommended to use this setting for clean-up operations that
526 shall be executed even when the service failed to start up
527 correctly. Commands configured with this setting need to be able to
528 operate even if the service failed starting up half-way and left
529 incompletely initialized data around. As the service's processes
530 have been terminated already when the commands specified with this
531 setting are executed they should not attempt to communicate with
532 them.
533 Note that all commands that are configured with this setting are
535 invoked with the result code of the service, as well as the main
536 process' exit code and status, set in the $SERVICE_RESULT,
537 $EXIT_CODE and $EXIT_STATUS environment variables, see
538 systemd.exec(5) for details.
539 RestartSec=
541 Configures the time to sleep before restarting a service (as
542 configured with Restart=). Takes a unit-less value in seconds, or a
543 time span value such as "5min 20s". Defaults to 100ms.
544 TimeoutStartSec=
546 Configures the time to wait for start-up. If a daemon service does
547 not signal start-up completion within the configured time, the
548 service will be considered failed and will be shut down again.
549 Takes a unit-less value in seconds, or a time span value such as
550 "5min 20s". Pass "infinity" to disable the timeout logic. Defaults
551 to DefaultTimeoutStartSec= from the manager configuration file,
552 except when Type=oneshot is used, in which case the timeout is
553 disabled by default (see systemd-system.conf(5)).
554 If a service of Type=notify sends "EXTEND_TIMEOUT_USEC=...", this
556 may cause the start time to be extended beyond TimeoutStartSec=.
557 The first receipt of this message must occur before
558 TimeoutStartSec= is exceeded, and once the start time has exended
559 beyond TimeoutStartSec=, the service manager will allow the service
560 to continue to start, provided the service repeats
561 "EXTEND_TIMEOUT_USEC=..." within the interval specified until the
562 service startup status is finished by "READY=1". (see
563 sd_notify(3)).
564 TimeoutStopSec=
566 This option serves two purposes. First, it configures the time to
567 wait for each ExecStop= command. If any of them times out,
568 subsequent ExecStop= commands are skipped and the service will be
569 terminated by SIGTERM. If no ExecStop= commands are specified, the
570 service gets the SIGTERM immediately. Second, it configures the
571 time to wait for the service itself to stop. If it doesn't
572 terminate in the specified time, it will be forcibly terminated by
573 SIGKILL (see KillMode= in systemd.kill(5)). Takes a unit-less value
574 in seconds, or a time span value such as "5min 20s". Pass
575 "infinity" to disable the timeout logic. Defaults to
576 DefaultTimeoutStopSec= from the manager configuration file (see
577 systemd-system.conf(5)).
578 If a service of Type=notify sends "EXTEND_TIMEOUT_USEC=...", this
580 may cause the stop time to be extended beyond TimeoutStopSec=. The
581 first receipt of this message must occur before TimeoutStopSec= is
582 exceeded, and once the stop time has exended beyond
583 TimeoutStopSec=, the service manager will allow the service to
584 continue to stop, provided the service repeats
585 "EXTEND_TIMEOUT_USEC=..." within the interval specified, or
586 terminates itself (see sd_notify(3)).
587 TimeoutAbortSec=
589 This option configures the time to wait for the service to
590 terminate when it was aborted due to a watchdog timeout (see
591 WatchdogSec=). If the service has a short TimeoutStopSec= this
592 option can be used to give the system more time to write a core
593 dump of the service. Upon expiration the service will be forcibly
594 terminated by SIGKILL (see KillMode= in systemd.kill(5)). The core
595 file will be truncated in this case. Use TimeoutAbortSec= to set a
596 sensible timeout for the core dumping per service that is large
597 enough to write all expected data while also being short enough to
598 handle the service failure in due time.
599 Takes a unit-less value in seconds, or a time span value such as
601 "5min 20s". Pass an empty value to skip the dedicated watchdog
602 abort timeout handling and fall back TimeoutStopSec=. Pass
603 "infinity" to disable the timeout logic. Defaults to
604 DefaultTimeoutAbortSec= from the manager configuration file (see
605 systemd-system.conf(5)).
606 If a service of Type=notify handles SIGABRT itself (instead of
608 relying on the kernel to write a core dump) it can send
609 "EXTEND_TIMEOUT_USEC=..." to extended the abort time beyond
610 TimeoutAbortSec=. The first receipt of this message must occur
611 before TimeoutAbortSec= is exceeded, and once the abort time has
612 exended beyond TimeoutAbortSec=, the service manager will allow the
613 service to continue to abort, provided the service repeats
614 "EXTEND_TIMEOUT_USEC=..." within the interval specified, or
615 terminates itself (see sd_notify(3)).
616 TimeoutSec=
618 A shorthand for configuring both TimeoutStartSec= and
619 TimeoutStopSec= to the specified value.
620 RuntimeMaxSec=
622 Configures a maximum time for the service to run. If this is used
623 and the service has been active for longer than the specified time
624 it is terminated and put into a failure state. Note that this
625 setting does not have any effect on Type=oneshot services, as they
626 terminate immediately after activation completed. Pass "infinity"
627 (the default) to configure no runtime limit.
628 If a service of Type=notify sends "EXTEND_TIMEOUT_USEC=...", this
630 may cause the runtime to be extended beyond RuntimeMaxSec=. The
631 first receipt of this message must occur before RuntimeMaxSec= is
632 exceeded, and once the runtime has exended beyond RuntimeMaxSec=,
633 the service manager will allow the service to continue to run,
634 provided the service repeats "EXTEND_TIMEOUT_USEC=..." within the
635 interval specified until the service shutdown is achieved by
636 "STOPPING=1" (or termination). (see sd_notify(3)).
637 WatchdogSec=
639 Configures the watchdog timeout for a service. The watchdog is
640 activated when the start-up is completed. The service must call
641 sd_notify(3) regularly with "WATCHDOG=1" (i.e. the "keep-alive
642 ping"). If the time between two such calls is larger than the
643 configured time, then the service is placed in a failed state and
644 it will be terminated with SIGABRT (or the signal specified by
645 WatchdogSignal=). By setting Restart= to on-failure, on-watchdog,
646 on-abnormal or always, the service will be automatically restarted.
647 The time configured here will be passed to the executed service
648 process in the WATCHDOG_USEC= environment variable. This allows
649 daemons to automatically enable the keep-alive pinging logic if
650 watchdog support is enabled for the service. If this option is
651 used, NotifyAccess= (see below) should be set to open access to the
652 notification socket provided by systemd. If NotifyAccess= is not
653 set, it will be implicitly set to main. Defaults to 0, which
654 disables this feature. The service can check whether the service
655 manager expects watchdog keep-alive notifications. See
656 sd_watchdog_enabled(3) for details. sd_event_set_watchdog(3) may
657 be used to enable automatic watchdog notification support.
658 Restart=
660 Configures whether the service shall be restarted when the service
661 process exits, is killed, or a timeout is reached. The service
662 process may be the main service process, but it may also be one of
663 the processes specified with ExecStartPre=, ExecStartPost=,
664 ExecStop=, ExecStopPost=, or ExecReload=. When the death of the
665 process is a result of systemd operation (e.g. service stop or
666 restart), the service will not be restarted. Timeouts include
667 missing the watchdog "keep-alive ping" deadline and a service
668 start, reload, and stop operation timeouts.
669 Takes one of no, on-success, on-failure, on-abnormal, on-watchdog,
671 on-abort, or always. If set to no (the default), the service will
672 not be restarted. If set to on-success, it will be restarted only
673 when the service process exits cleanly. In this context, a clean
674 exit means an exit code of 0, or one of the signals SIGHUP, SIGINT,
675 SIGTERM or SIGPIPE, and additionally, exit statuses and signals
676 specified in SuccessExitStatus=. If set to on-failure, the service
677 will be restarted when the process exits with a non-zero exit code,
678 is terminated by a signal (including on core dump, but excluding
679 the aforementioned four signals), when an operation (such as
680 service reload) times out, and when the configured watchdog timeout
681 is triggered. If set to on-abnormal, the service will be restarted
682 when the process is terminated by a signal (including on core dump,
683 excluding the aforementioned four signals), when an operation times
684 out, or when the watchdog timeout is triggered. If set to on-abort,
685 the service will be restarted only if the service process exits due
686 to an uncaught signal not specified as a clean exit status. If set
687 to on-watchdog, the service will be restarted only if the watchdog
688 timeout for the service expires. If set to always, the service will
689 be restarted regardless of whether it exited cleanly or not, got
690 terminated abnormally by a signal, or hit a timeout.
691 Table 2. Exit causes and the effect of the Restart= settings on
693 them
694 ┌──────────────┬────┬────────┬────────────┬────────────┬─────────────┬──────────┬─────────────┐
695 │Restart │ no │ always │ on-success │ on-failure │ on-abnormal │ on-abort │ on-watchdog │
696 │settings/Exit │ │ │ │ │ │ │ │
697 │causes │ │ │ │ │ │ │ │
698 ├──────────────┼────┼────────┼────────────┼────────────┼─────────────┼──────────┼─────────────┤
699 │Clean exit │ │ X │ X │ │ │ │ │
700 │code or │ │ │ │ │ │ │ │
701 │signal │ │ │ │ │ │ │ │
702 ├──────────────┼────┼────────┼────────────┼────────────┼─────────────┼──────────┼─────────────┤
703 │Unclean exit │ │ X │ │ X │ │ │ │
704 │code │ │ │ │ │ │ │ │
705 ├──────────────┼────┼────────┼────────────┼────────────┼─────────────┼──────────┼─────────────┤
706 │Unclean │ │ X │ │ X │ X │ X │ │
707 │signal │ │ │ │ │ │ │ │
708 ├──────────────┼────┼────────┼────────────┼────────────┼─────────────┼──────────┼─────────────┤
709 │Timeout │ │ X │ │ X │ X │ │ │
710 ├──────────────┼────┼────────┼────────────┼────────────┼─────────────┼──────────┼─────────────┤
711 │Watchdog │ │ X │ │ X │ X │ │ X │
712 └──────────────┴────┴────────┴────────────┴────────────┴─────────────┴──────────┴─────────────┘
713 As exceptions to the setting above, the service will not be
714 restarted if the exit code or signal is specified in
715 RestartPreventExitStatus= (see below) or the service is stopped
716 with systemctl stop or an equivalent operation. Also, the services
717 will always be restarted if the exit code or signal is specified in
718 RestartForceExitStatus= (see below).
719 Note that service restart is subject to unit start rate limiting
721 configured with StartLimitIntervalSec= and StartLimitBurst=, see
722 systemd.unit(5) for details. A restarted service enters the failed
723 state only after the start limits are reached.
724 Setting this to on-failure is the recommended choice for
726 long-running services, in order to increase reliability by
727 attempting automatic recovery from errors. For services that shall
728 be able to terminate on their own choice (and avoid immediate
729 restarting), on-abnormal is an alternative choice.
730 SuccessExitStatus=
732 Takes a list of exit status definitions that, when returned by the
733 main service process, will be considered successful termination, in
734 addition to the normal successful exit code 0 and the signals
735 SIGHUP, SIGINT, SIGTERM, and SIGPIPE. Exit status definitions can
736 be numeric exit codes, termination code names, or termination
737 signal names, separated by spaces. See the Process Exit Codes
738 section in systemd.exec(5) for a list of termination codes names
739 (for this setting only the part without the "EXIT_" or "EX_" prefix
740 should be used). See signal(7) for a list of signal names.
741 This option may appear more than once, in which case the list of
743 successful exit statuses is merged. If the empty string is assigned
744 to this option, the list is reset, all prior assignments of this
745 option will have no effect.
746 Example 1. A service with with the SuccessExitStatus= setting
748 SuccessExitStatus=TEMPFAIL 250 SIGUSR1
750 Exit codes 75 (TEMPFAIL), 250, and the termination signal SIGKILL
752 are considered clean service terminations.
753 Note: systemd-analyze exit-codes may be used to list exit codes and
755 translate between numerical code values and names.
756 RestartPreventExitStatus=
758 Takes a list of exit status definitions that, when returned by the
759 main service process, will prevent automatic service restarts,
760 regardless of the restart setting configured with Restart=. Exit
761 status definitions can either be numeric exit codes or termination
762 signal names, and are separated by spaces. Defaults to the empty
763 list, so that, by default, no exit status is excluded from the
764 configured restart logic. For example:
765 RestartPreventExitStatus=1 6 SIGABRT
767 ensures that exit codes 1 and 6 and the termination signal SIGABRT
769 will not result in automatic service restarting. This option may
770 appear more than once, in which case the list of restart-preventing
771 statuses is merged. If the empty string is assigned to this option,
772 the list is reset and all prior assignments of this option will
773 have no effect.
774 Note that this setting has no effect on processes configured via
776 ExecStartPre=, ExecStartPost=, ExecStop=, ExecStopPost= or
777 ExecReload=, but only on the main service process, i.e. either the
778 one invoked by ExecStart= or (depending on Type=, PIDFile=, ...)
779 the otherwise configured main process.
780 RestartForceExitStatus=
782 Takes a list of exit status definitions that, when returned by the
783 main service process, will force automatic service restarts,
784 regardless of the restart setting configured with Restart=. The
785 argument format is similar to RestartPreventExitStatus=.
786 RootDirectoryStartOnly=
788 Takes a boolean argument. If true, the root directory, as
789 configured with the RootDirectory= option (see systemd.exec(5) for
790 more information), is only applied to the process started with
791 ExecStart=, and not to the various other ExecStartPre=,
792 ExecStartPost=, ExecReload=, ExecStop=, and ExecStopPost= commands.
793 If false, the setting is applied to all configured commands the
794 same way. Defaults to false.
795 NonBlocking=
797 Set the O_NONBLOCK flag for all file descriptors passed via
798 socket-based activation. If true, all file descriptors >= 3 (i.e.
799 all except stdin, stdout, stderr), excluding those passed in via
800 the file descriptor storage logic (see FileDescriptorStoreMax= for
801 details), will have the O_NONBLOCK flag set and hence are in
802 non-blocking mode. This option is only useful in conjunction with a
803 socket unit, as described in systemd.socket(5) and has no effect on
804 file descriptors which were previously saved in the file-descriptor
805 store for example. Defaults to false.
806 NotifyAccess=
808 Controls access to the service status notification socket, as
809 accessible via the sd_notify(3) call. Takes one of none (the
810 default), main, exec or all. If none, no daemon status updates are
811 accepted from the service processes, all status update messages are
812 ignored. If main, only service updates sent from the main process
813 of the service are accepted. If exec, only service updates sent
814 from any of the main or control processes originating from one of
815 the Exec*= commands are accepted. If all, all services updates from
816 all members of the service's control group are accepted. This
817 option should be set to open access to the notification socket when
818 using Type=notify or WatchdogSec= (see above). If those options are
819 used but NotifyAccess= is not configured, it will be implicitly set
820 to main.
821 Note that sd_notify() notifications may be attributed to units
823 correctly only if either the sending process is still around at the
824 time PID 1 processes the message, or if the sending process is
825 explicitly runtime-tracked by the service manager. The latter is
826 the case if the service manager originally forked off the process,
827 i.e. on all processes that match main or exec. Conversely, if an
828 auxiliary process of the unit sends an sd_notify() message and
829 immediately exits, the service manager might not be able to
830 properly attribute the message to the unit, and thus will ignore
831 it, even if NotifyAccess=all is set for it.
832 Sockets=
834 Specifies the name of the socket units this service shall inherit
835 socket file descriptors from when the service is started. Normally,
836 it should not be necessary to use this setting, as all socket file
837 descriptors whose unit shares the same name as the service (subject
838 to the different unit name suffix of course) are passed to the
839 spawned process.
840 Note that the same socket file descriptors may be passed to
842 multiple processes simultaneously. Also note that a different
843 service may be activated on incoming socket traffic than the one
844 which is ultimately configured to inherit the socket file
845 descriptors. Or, in other words: the Service= setting of .socket
846 units does not have to match the inverse of the Sockets= setting of
847 the .service it refers to.
848 This option may appear more than once, in which case the list of
850 socket units is merged. Note that once set, clearing the list of
851 sockets again (for example, by assigning the empty string to this
852 option) is not supported.
853 FileDescriptorStoreMax=
855 Configure how many file descriptors may be stored in the service
856 manager for the service using sd_pid_notify_with_fds(3)'s
857 "FDSTORE=1" messages. This is useful for implementing services that
858 can restart after an explicit request or a crash without losing
859 state. Any open sockets and other file descriptors which should not
860 be closed during the restart may be stored this way. Application
861 state can either be serialized to a file in /run, or better, stored
862 in a memfd_create(2) memory file descriptor. Defaults to 0, i.e. no
863 file descriptors may be stored in the service manager. All file
864 descriptors passed to the service manager from a specific service
865 are passed back to the service's main process on the next service
866 restart. Any file descriptors passed to the service manager are
867 automatically closed when POLLHUP or POLLERR is seen on them, or
868 when the service is fully stopped and no job is queued or being
869 executed for it. If this option is used, NotifyAccess= (see above)
870 should be set to open access to the notification socket provided by
871 systemd. If NotifyAccess= is not set, it will be implicitly set to
872 main.
873 USBFunctionDescriptors=
875 Configure the location of a file containing USB FunctionFS[2]
876 descriptors, for implementation of USB gadget functions. This is
877 used only in conjunction with a socket unit with ListenUSBFunction=
878 configured. The contents of this file are written to the ep0 file
879 after it is opened.
880 USBFunctionStrings=
882 Configure the location of a file containing USB FunctionFS strings.
883 Behavior is similar to USBFunctionDescriptors= above.
884 OOMPolicy=
886 Configure the Out-Of-Memory (OOM) killer policy. On Linux, when
887 memory becomes scarce the kernel might decide to kill a running
888 process in order to free up memory and reduce memory pressure. This
889 setting takes one of continue, stop or kill. If set to continue and
890 a process of the service is killed by the kernel's OOM killer this
891 is logged but the service continues running. If set to stop the
892 event is logged but the service is terminated cleanly by the
893 service manager. If set to kill and one of the service's processes
894 is killed by the OOM killer the kernel is instructed to kill all
895 remaining processes of the service, too. Defaults to the setting
896 DefaultOOMPolicy= in system.conf(5) is set to, except for services
897 where Delegate= is turned on, where it defaults to continue.
898 Use the OOMScoreAdjust= setting to configure whether processes of
900 the unit shall be considered preferred or less preferred candidates
901 for process termination by the Linux OOM killer logic. See
902 systemd.exec(5) for details.
903 Check systemd.exec(5) and systemd.kill(5) for more settings.
905
907 This section describes command line parsing and variable and specifier
908 substitutions for ExecStart=, ExecStartPre=, ExecStartPost=,
909 ExecReload=, ExecStop=, and ExecStopPost= options.
910 Multiple command lines may be concatenated in a single directive by
912 separating them with semicolons (these semicolons must be passed as
913 separate words). Lone semicolons may be escaped as "\;".
914 Each command line is split on whitespace, with the first item being the
916 command to execute, and the subsequent items being the arguments.
917 Double quotes ("...") and single quotes ('...') may be used to wrap a
918 whole item (the opening quote may appear only at the beginning or after
919 whitespace that is not quoted, and the closing quote must be followed
920 by whitespace or the end of line), in which case everything until the
921 next matching quote becomes part of the same argument. Quotes
922 themselves are removed. C-style escapes are also supported. The table
923 below contains the list of known escape patterns. Only escape patterns
924 which match the syntax in the table are allowed; other patterns may be
925 added in the future and unknown patterns will result in a warning. In
926 particular, any backslashes should be doubled. Finally, a trailing
927 backslash ("\") may be used to merge lines.
928 This syntax is inspired by shell syntax, but only the meta-characters
930 and expansions described in the following paragraphs are understood,
931 and the expansion of variables is different. Specifically, redirection
932 using "<", "<<", ">", and ">>", pipes using "|", running programs in
933 the background using "&", and other elements of shell syntax are not
934 supported.
935 The command to execute may contain spaces, but control characters are
937 not allowed.
938 The command line accepts "%" specifiers as described in
940 systemd.unit(5).
941 Basic environment variable substitution is supported. Use "${FOO}" as
943 part of a word, or as a word of its own, on the command line, in which
944 case it will be replaced by the value of the environment variable
945 including all whitespace it contains, resulting in a single argument.
946 Use "$FOO" as a separate word on the command line, in which case it
947 will be replaced by the value of the environment variable split at
948 whitespace, resulting in zero or more arguments. For this type of
949 expansion, quotes are respected when splitting into words, and
950 afterwards removed.
951 If the command is not a full (absolute) path, it will be resolved to a
953 full path using a fixed search path determinted at compilation time.
954 Searched directories include /usr/local/bin/, /usr/bin/, /bin/ on
955 systems using split /usr/bin/ and /bin/ directories, and their sbin/
956 counterparts on systems using split bin/ and sbin/. It is thus safe to
957 use just the executable name in case of executables located in any of
958 the "standard" directories, and an absolute path must be used in other
959 cases. Using an absolute path is recommended to avoid ambiguity. Hint:
960 this search path may be queried using systemd-path
961 search-binaries-default.
962 Example:
964 Environment="ONE=one" 'TWO=two two'
966 ExecStart=echo $ONE $TWO ${TWO}
967 This will execute /bin/echo with four arguments: "one", "two", "two",
969 and "two two".
970 Example:
972 Environment=ONE='one' "TWO='two two' too" THREE=
974 ExecStart=/bin/echo ${ONE} ${TWO} ${THREE}
975 ExecStart=/bin/echo $ONE $TWO $THREE
976 This results in /bin/echo being called twice, the first time with
978 arguments "'one'", "'two two' too", "", and the second time with
979 arguments "one", "two two", "too".
980 To pass a literal dollar sign, use "$$". Variables whose value is not
982 known at expansion time are treated as empty strings. Note that the
983 first argument (i.e. the program to execute) may not be a variable.
984 Variables to be used in this fashion may be defined through
986 Environment= and EnvironmentFile=. In addition, variables listed in the
987 section "Environment variables in spawned processes" in
988 systemd.exec(5), which are considered "static configuration", may be
989 used (this includes e.g. $USER, but not $TERM).
990 Note that shell command lines are not directly supported. If shell
992 command lines are to be used, they need to be passed explicitly to a
993 shell implementation of some kind. Example:
994 ExecStart=sh -c 'dmesg | tac'
996 Example:
998 ExecStart=echo one ; echo "two two"
1000 This will execute echo two times, each time with one argument: "one"
1002 and "two two", respectively. Because two commands are specified,
1003 Type=oneshot must be used.
1004 Example:
1006 ExecStart=echo / >/dev/null & \; \
1008 ls
1009 This will execute echo with five arguments: "/", ">/dev/null", "&",
1011 ";", and "ls".
1012 Table 3. C escapes supported in command lines and environment variables
1014 ┌────────┬─────────────────────────┐
1015 │Literal │ Actual value │
1016 ├────────┼─────────────────────────┤
1017 │"\a" │ bell │
1018 ├────────┼─────────────────────────┤
1019 │"\b" │ backspace │
1020 ├────────┼─────────────────────────┤
1021 │"\f" │ form feed │
1022 ├────────┼─────────────────────────┤
1023 │"\n" │ newline │
1024 ├────────┼─────────────────────────┤
1025 │"\r" │ carriage return │
1026 ├────────┼─────────────────────────┤
1027 │"\t" │ tab │
1028 ├────────┼─────────────────────────┤
1029 │"\v" │ vertical tab │
1030 ├────────┼─────────────────────────┤
1031 │"\\" │ backslash │
1032 ├────────┼─────────────────────────┤
1033 │"\"" │ double quotation mark │
1034 ├────────┼─────────────────────────┤
1035 │"\'" │ single quotation mark │
1036 ├────────┼─────────────────────────┤
1037 │"\s" │ space │
1038 ├────────┼─────────────────────────┤
1039 │"\xxx" │ character number xx in │
1040 │ │ hexadecimal encoding │
1041 ├────────┼─────────────────────────┤
1042 │"\nnn" │ character number nnn in │
1043 │ │ octal encoding │
1044 └────────┴─────────────────────────┘
1045
1047 Example 2. Simple service
1048 The following unit file creates a service that will execute
1050 /usr/sbin/foo-daemon. Since no Type= is specified, the default
1051 Type=simple will be assumed. systemd will assume the unit to be started
1052 immediately after the program has begun executing.
1053 [Unit]
1055 Description=Foo
1056 [Service]
1058 ExecStart=/usr/sbin/foo-daemon
1059 [Install]
1061 WantedBy=multi-user.target
1062 Note that systemd assumes here that the process started by systemd will
1064 continue running until the service terminates. If the program
1065 daemonizes itself (i.e. forks), please use Type=forking instead.
1066 Since no ExecStop= was specified, systemd will send SIGTERM to all
1068 processes started from this service, and after a timeout also SIGKILL.
1069 This behavior can be modified, see systemd.kill(5) for details.
1070 Note that this unit type does not include any type of notification when
1072 a service has completed initialization. For this, you should use other
1073 unit types, such as Type=notify if the service understands systemd's
1074 notification protocol, Type=forking if the service can background
1075 itself or Type=dbus if the unit acquires a DBus name once
1076 initialization is complete. See below.
1077 Example 3. Oneshot service
1079 Sometimes, units should just execute an action without keeping active
1081 processes, such as a filesystem check or a cleanup action on boot. For
1082 this, Type=oneshot exists. Units of this type will wait until the
1083 process specified terminates and then fall back to being inactive. The
1084 following unit will perform a cleanup action:
1085 [Unit]
1087 Description=Cleanup old Foo data
1088 [Service]
1090 Type=oneshot
1091 ExecStart=/usr/sbin/foo-cleanup
1092 [Install]
1094 WantedBy=multi-user.target
1095 Note that systemd will consider the unit to be in the state "starting"
1097 until the program has terminated, so ordered dependencies will wait for
1098 the program to finish before starting themselves. The unit will revert
1099 to the "inactive" state after the execution is done, never reaching the
1100 "active" state. That means another request to start the unit will
1101 perform the action again.
1102 Type=oneshot are the only service units that may have more than one
1104 ExecStart= specified. They will be executed in order until either they
1105 are all successful or one of them fails.
1106 Example 4. Stoppable oneshot service
1108 Similarly to the oneshot services, there are sometimes units that need
1110 to execute a program to set up something and then execute another to
1111 shut it down, but no process remains active while they are considered
1112 "started". Network configuration can sometimes fall into this category.
1113 Another use case is if a oneshot service shall not be executed each
1114 time when they are pulled in as a dependency, but only the first time.
1115 For this, systemd knows the setting RemainAfterExit=yes, which causes
1117 systemd to consider the unit to be active if the start action exited
1118 successfully. This directive can be used with all types, but is most
1119 useful with Type=oneshot and Type=simple. With Type=oneshot, systemd
1120 waits until the start action has completed before it considers the unit
1121 to be active, so dependencies start only after the start action has
1122 succeeded. With Type=simple, dependencies will start immediately after
1123 the start action has been dispatched. The following unit provides an
1124 example for a simple static firewall.
1125 [Unit]
1127 Description=Simple firewall
1128 [Service]
1130 Type=oneshot
1131 RemainAfterExit=yes
1132 ExecStart=/usr/local/sbin/simple-firewall-start
1133 ExecStop=/usr/local/sbin/simple-firewall-stop
1134 [Install]
1136 WantedBy=multi-user.target
1137 Since the unit is considered to be running after the start action has
1139 exited, invoking systemctl start on that unit again will cause no
1140 action to be taken.
1141 Example 5. Traditional forking services
1143 Many traditional daemons/services background (i.e. fork, daemonize)
1145 themselves when starting. Set Type=forking in the service's unit file
1146 to support this mode of operation. systemd will consider the service to
1147 be in the process of initialization while the original program is still
1148 running. Once it exits successfully and at least a process remains (and
1149 RemainAfterExit=no), the service is considered started.
1150 Often, a traditional daemon only consists of one process. Therefore, if
1152 only one process is left after the original process terminates, systemd
1153 will consider that process the main process of the service. In that
1154 case, the $MAINPID variable will be available in ExecReload=,
1155 ExecStop=, etc.
1156 In case more than one process remains, systemd will be unable to
1158 determine the main process, so it will not assume there is one. In that
1159 case, $MAINPID will not expand to anything. However, if the process
1160 decides to write a traditional PID file, systemd will be able to read
1161 the main PID from there. Please set PIDFile= accordingly. Note that the
1162 daemon should write that file before finishing with its initialization.
1163 Otherwise, systemd might try to read the file before it exists.
1164 The following example shows a simple daemon that forks and just starts
1166 one process in the background:
1167 [Unit]
1169 Description=Some simple daemon
1170 [Service]
1172 Type=forking
1173 ExecStart=/usr/sbin/my-simple-daemon -d
1174 [Install]
1176 WantedBy=multi-user.target
1177 Please see systemd.kill(5) for details on how you can influence the way
1179 systemd terminates the service.
1180 Example 6. DBus services
1182 For services that acquire a name on the DBus system bus, use Type=dbus
1184 and set BusName= accordingly. The service should not fork (daemonize).
1185 systemd will consider the service to be initialized once the name has
1186 been acquired on the system bus. The following example shows a typical
1187 DBus service:
1188 [Unit]
1190 Description=Simple DBus service
1191 [Service]
1193 Type=dbus
1194 BusName=org.example.simple-dbus-service
1195 ExecStart=/usr/sbin/simple-dbus-service
1196 [Install]
1198 WantedBy=multi-user.target
1199 For bus-activatable services, do not include a "[Install]" section in
1201 the systemd service file, but use the SystemdService= option in the
1202 corresponding DBus service file, for example
1203 (/usr/share/dbus-1/system-services/org.example.simple-dbus-service.service):
1204 [D-BUS Service]
1206 Name=org.example.simple-dbus-service
1207 Exec=/usr/sbin/simple-dbus-service
1208 User=root
1209 SystemdService=simple-dbus-service.service
1210 Please see systemd.kill(5) for details on how you can influence the way
1212 systemd terminates the service.
1213 Example 7. Services that notify systemd about their initialization
1215 Type=simple services are really easy to write, but have the major
1217 disadvantage of systemd not being able to tell when initialization of
1218 the given service is complete. For this reason, systemd supports a
1219 simple notification protocol that allows daemons to make systemd aware
1220 that they are done initializing. Use Type=notify for this. A typical
1221 service file for such a daemon would look like this:
1222 [Unit]
1224 Description=Simple notifying service
1225 [Service]
1227 Type=notify
1228 ExecStart=/usr/sbin/simple-notifying-service
1229 [Install]
1231 WantedBy=multi-user.target
1232 Note that the daemon has to support systemd's notification protocol,
1234 else systemd will think the service has not started yet and kill it
1235 after a timeout. For an example of how to update daemons to support
1236 this protocol transparently, take a look at sd_notify(3). systemd will
1237 consider the unit to be in the 'starting' state until a readiness
1238 notification has arrived.
1239 Please see systemd.kill(5) for details on how you can influence the way
1241 systemd terminates the service.
1242
1244 systemd(1), systemctl(1), systemd-system.conf(5), systemd.unit(5),
1245 systemd.exec(5), systemd.resource-control(5), systemd.kill(5),
1246 systemd.directives(7), systemd-run(1)
1247
1249 1. Incompatibilities with SysV
1250 https://www.freedesktop.org/wiki/Software/systemd/Incompatibilities
1251 2. USB FunctionFS
1253 https://www.kernel.org/doc/Documentation/usb/functionfs.txt
1254systemd 243 SYSTEMD.SERVICE(5)