]> git.ipfire.org Git - thirdparty/systemd.git/blob - man/bootup.xml
verify: use manager_load_startable_unit_or_warn() to load units for verification
[thirdparty/systemd.git] / man / bootup.xml
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">
4
5 <!--
6 SPDX-License-Identifier: LGPL-2.1+
7
8 This file is part of systemd.
9
10 Copyright 2012 Lennart Poettering
11
12 systemd is free software; you can redistribute it and/or modify it
13 under the terms of the GNU Lesser General Public License as published by
14 the Free Software Foundation; either version 2.1 of the License, or
15 (at your option) any later version.
16
17 systemd is distributed in the hope that it will be useful, but
18 WITHOUT ANY WARRANTY; without even the implied warranty of
19 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
20 Lesser General Public License for more details.
21
22 You should have received a copy of the GNU Lesser General Public License
23 along with systemd; If not, see <http://www.gnu.org/licenses/>.
24 -->
25
26 <refentry id="bootup">
27
28 <refentryinfo>
29 <title>bootup</title>
30 <productname>systemd</productname>
31
32 <authorgroup>
33 <author>
34 <contrib>Developer</contrib>
35 <firstname>Lennart</firstname>
36 <surname>Poettering</surname>
37 <email>lennart@poettering.net</email>
38 </author>
39 </authorgroup>
40 </refentryinfo>
41
42 <refmeta>
43 <refentrytitle>bootup</refentrytitle>
44 <manvolnum>7</manvolnum>
45 </refmeta>
46
47 <refnamediv>
48 <refname>bootup</refname>
49 <refpurpose>System bootup process</refpurpose>
50 </refnamediv>
51
52 <refsect1>
53 <title>Description</title>
54
55 <para>A number of different components are involved in the system
56 boot. Immediately after power-up, the system BIOS will do minimal
57 hardware initialization, and hand control over to a boot loader
58 stored on a persistent storage device. This boot loader will then
59 invoke an OS kernel from disk (or the network). In the Linux case,
60 this kernel (optionally) extracts and executes an initial RAM disk
61 image (initrd), such as generated by
62 <citerefentry project='die-net'><refentrytitle>dracut</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
63 which looks for the root file system (possibly using
64 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>
65 for this). After the root file system is found and mounted, the
66 initrd hands over control to the host's system manager (such as
67 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>)
68 stored on the OS image, which is then responsible for probing all
69 remaining hardware, mounting all necessary file systems and
70 spawning all configured services.</para>
71
72 <para>On shutdown, the system manager stops all services, unmounts
73 all file systems (detaching the storage technologies backing
74 them), and then (optionally) jumps back into the initrd code which
75 unmounts/detaches the root file system and the storage it resides
76 on. As a last step, the system is powered down.</para>
77
78 <para>Additional information about the system boot process may be
79 found in
80 <citerefentry project='man-pages'><refentrytitle>boot</refentrytitle><manvolnum>7</manvolnum></citerefentry>.</para>
81 </refsect1>
82
83 <refsect1>
84 <title>System Manager Bootup</title>
85
86 <para>At boot, the system manager on the OS image is responsible
87 for initializing the required file systems, services and drivers
88 that are necessary for operation of the system. On
89 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>
90 systems, this process is split up in various discrete steps which
91 are exposed as target units. (See
92 <citerefentry><refentrytitle>systemd.target</refentrytitle><manvolnum>5</manvolnum></citerefentry>
93 for detailed information about target units.) The boot-up process
94 is highly parallelized so that the order in which specific target
95 units are reached is not deterministic, but still adheres to a
96 limited amount of ordering structure.</para>
97
98 <para>When systemd starts up the system, it will activate all
99 units that are dependencies of <filename>default.target</filename>
100 (as well as recursively all dependencies of these dependencies).
101 Usually, <filename>default.target</filename> is simply an alias of
102 <filename>graphical.target</filename> or
103 <filename>multi-user.target</filename>, depending on whether the
104 system is configured for a graphical UI or only for a text
105 console. To enforce minimal ordering between the units pulled in,
106 a number of well-known target units are available, as listed on
107 <citerefentry><refentrytitle>systemd.special</refentrytitle><manvolnum>7</manvolnum></citerefentry>.</para>
108
109 <para>The following chart is a structural overview of these
110 well-known units and their position in the boot-up logic. The
111 arrows describe which units are pulled in and ordered before which
112 other units. Units near the top are started before units nearer to
113 the bottom of the chart.</para>
114
115 <!-- note: do not use unicode ellipsis here, because docbook will replace that
116 with three dots anyway, messing up alignment -->
117 <programlisting>local-fs-pre.target
118 |
119 v
120 (various mounts and (various swap (various cryptsetup
121 fsck services...) devices...) devices...) (various low-level (various low-level
122 | | | services: udevd, API VFS mounts:
123 v v v tmpfiles, random mqueue, configfs,
124 local-fs.target swap.target cryptsetup.target seed, sysctl, ...) debugfs, ...)
125 | | | | |
126 \__________________|_________________ | ___________________|____________________/
127 \|/
128 v
129 sysinit.target
130 |
131 ____________________________________/|\________________________________________
132 / | | | \
133 | | | | |
134 v v | v v
135 (various (various | (various rescue.service
136 timers...) paths...) | sockets...) |
137 | | | | v
138 v v | v <emphasis>rescue.target</emphasis>
139 timers.target paths.target | sockets.target
140 | | | |
141 v \_________________ | ___________________/
142 \|/
143 v
144 basic.target
145 |
146 ____________________________________/| emergency.service
147 / | | |
148 | | | v
149 v v v <emphasis>emergency.target</emphasis>
150 display- (various system (various system
151 manager.service services services)
152 | required for |
153 | graphical UIs) v
154 | | <emphasis>multi-user.target</emphasis>
155 | | |
156 \_________________ | _________________/
157 \|/
158 v
159 <emphasis>graphical.target</emphasis></programlisting>
160
161 <para>Target units that are commonly used as boot targets are
162 <emphasis>emphasized</emphasis>. These units are good choices as
163 goal targets, for example by passing them to the
164 <varname>systemd.unit=</varname> kernel command line option (see
165 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>)
166 or by symlinking <filename>default.target</filename> to them.
167 </para>
168
169 <para><filename>timers.target</filename> is pulled-in by
170 <filename>basic.target</filename> asynchronously. This allows
171 timers units to depend on services which become only available
172 later in boot.</para>
173 </refsect1>
174
175 <refsect1>
176 <title>Bootup in the Initial RAM Disk (initrd)</title>
177 <para>The initial RAM disk implementation (initrd) can be set up
178 using systemd as well. In this case, boot up inside the initrd
179 follows the following structure.</para>
180
181 <para>The default target in the initrd is
182 <filename>initrd.target</filename>. The bootup process begins
183 identical to the system manager bootup (see above) until it
184 reaches <filename>basic.target</filename>. From there, systemd
185 approaches the special target <filename>initrd.target</filename>.
186
187 Before any file systems are mounted, it must be determined whether
188 the system will resume from hibernation or proceed with normal boot.
189 This is accomplished by <filename>systemd-hibernate-resume@.service</filename>
190 which must be finished before <filename>local-fs-pre.target</filename>,
191 so no filesystems can be mounted before the check is complete.
192
193 When the root device becomes available,
194 <filename>initd-root-device.target</filename> is reached.
195 If the root device can be mounted at
196 <filename>/sysroot</filename>, the
197 <filename>sysroot.mount</filename> unit becomes active and
198 <filename>initrd-root-fs.target</filename> is reached. The service
199 <filename>initrd-parse-etc.service</filename> scans
200 <filename>/sysroot/etc/fstab</filename> for a possible
201 <filename>/usr</filename> mount point and additional entries
202 marked with the <emphasis>x-initrd.mount</emphasis> option. All
203 entries found are mounted below <filename>/sysroot</filename>, and
204 <filename>initrd-fs.target</filename> is reached. The service
205 <filename>initrd-cleanup.service</filename> isolates to the
206 <filename>initrd-switch-root.target</filename>, where cleanup
207 services can run. As the very last step, the
208 <filename>initrd-switch-root.service</filename> is activated,
209 which will cause the system to switch its root to
210 <filename>/sysroot</filename>.
211 </para>
212
213 <programlisting> : (beginning identical to above)
214 :
215 v
216 basic.target
217 | emergency.service
218 ______________________/| |
219 / | v
220 | initrd-root-device.target <emphasis>emergency.target</emphasis>
221 | |
222 | v
223 | sysroot.mount
224 | |
225 | v
226 | initrd-root-fs.target
227 | |
228 | v
229 v initrd-parse-etc.service
230 (custom initrd |
231 services...) v
232 | (sysroot-usr.mount and
233 | various mounts marked
234 | with fstab option
235 | x-initrd.mount...)
236 | |
237 | v
238 | initrd-fs.target
239 \______________________ |
240 \|
241 v
242 initrd.target
243 |
244 v
245 initrd-cleanup.service
246 isolates to
247 initrd-switch-root.target
248 |
249 v
250 ______________________/|
251 / v
252 | initrd-udevadm-cleanup-db.service
253 v |
254 (custom initrd |
255 services...) |
256 \______________________ |
257 \|
258 v
259 initrd-switch-root.target
260 |
261 v
262 initrd-switch-root.service
263 |
264 v
265 Transition to Host OS</programlisting>
266 </refsect1>
267
268 <refsect1>
269 <title>System Manager Shutdown</title>
270
271 <para>System shutdown with systemd also consists of various target
272 units with some minimal ordering structure applied:</para>
273
274 <programlisting> (conflicts with (conflicts with
275 all system all file system
276 services) mounts, swaps,
277 | cryptsetup
278 | devices, ...)
279 | |
280 v v
281 shutdown.target umount.target
282 | |
283 \_______ ______/
284 \ /
285 v
286 (various low-level
287 services)
288 |
289 v
290 final.target
291 |
292 _____________________________________/ \_________________________________
293 / | | \
294 | | | |
295 v v v v
296 systemd-reboot.service systemd-poweroff.service systemd-halt.service systemd-kexec.service
297 | | | |
298 v v v v
299 <emphasis>reboot.target</emphasis> <emphasis>poweroff.target</emphasis> <emphasis>halt.target</emphasis> <emphasis>kexec.target</emphasis></programlisting>
300
301 <para>Commonly used system shutdown targets are <emphasis>emphasized</emphasis>.</para>
302
303 <para>Note that
304 <citerefentry><refentrytitle>systemd-halt.service</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
305 <filename>systemd-reboot.service</filename>, <filename>systemd-poweroff.service</filename> and
306 <filename>systemd-kexec.service</filename> will transition the system and server manager (PID 1) into the second
307 phase of system shutdown (implemented in the <filename>systemd-shutdown</filename> binary), which will unmount any
308 remaining file systems, kill any remaining processes and release any other remaining resources, in a simple and
309 robust fashion, without taking any service or unit concept into account anymore. At that point, regular
310 applications and resources are generally terminated and released already, the second phase hence operates only as
311 safety net for everything that couldn't be stopped or released for some reason during the primary, unit-based
312 shutdown phase described above.</para>
313 </refsect1>
314
315 <refsect1>
316 <title>See Also</title>
317 <para>
318 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
319 <citerefentry project='man-pages'><refentrytitle>boot</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
320 <citerefentry><refentrytitle>systemd.special</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
321 <citerefentry><refentrytitle>systemd.target</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
322 <citerefentry><refentrytitle>systemd-halt.service</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
323 <citerefentry project='die-net'><refentrytitle>dracut</refentrytitle><manvolnum>8</manvolnum></citerefentry>
324 </para>
325 </refsect1>
326
327 </refentry>