1 <?xml version="1.0" encoding="ISO-8859-1"?>
2 <!DOCTYPE sect1 PUBLIC "-//OASIS//DTD DocBook XML V4.5//EN"
3 "http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd" [
4 <!ENTITY % general-entities SYSTEM "../general.ent">
8 <sect1 id="ch-scripts-systemd-custom" revision="systemd">
9 <?dbhtml filename="systemd-custom.html"?>
11 <title>Systemd Usage and Configuration</title>
13 <indexterm zone="ch-scripts-systemd-custom">
14 <primary sortas="e-Systemd">Systemd Customization</primary>
18 <title>Basic Configuration</title>
20 <para>The <filename>/etc/systemd/system.conf</filename> file contains a set
21 of options to control basic systemd operations. The default file has all
22 entries commented out with the default settings indicated. This file is
23 where the log level may be changed as well as some basic logging settings.
24 See the <filename>systemd-system.conf(5)</filename> manual page for details
25 on each configuration option.</para>
30 <title>Disabling Screen Clearing at Boot Time</title>
32 <para>The normal behavior for systemd is to clear the screen at
33 the end of the boot sequence. If desired, this behavior may be
34 changed by running the following command:</para>
36 <screen role="nodump"><userinput>mkdir -pv /etc/systemd/system/getty@tty1.service.d
38 cat > /etc/systemd/system/getty@tty1.service.d/noclear.conf << EOF
40 TTYVTDisallocate=no</literal>
41 EOF</userinput></screen>
43 <para>The boot messages can always be reviewed by using the
44 <userinput>journalctl -b</userinput> command as the root user.</para>
49 <title>Disabling tmpfs for /tmp</title>
51 <para>By default, <filename class="directory">/tmp</filename> is created as
52 a tmpfs. If this is not desired, it can be overridden by the following:</para>
54 <screen role="nodump"><userinput>ln -sfv /dev/null /etc/systemd/system/tmp.mount</userinput></screen>
56 <para>Alternatively, if a a separate partition for
57 <filename class="directory">/tmp</filename> is desired, specify that
58 partition in an <filename>/etc/fstab</filename> entry.</para>
62 Do not create the symbolic link above if a separate partition is used
63 for <filename class="directory">/tmp</filename>. This will prvent the
64 root file system (/) from being remounted r/w and make the system
72 <title>Configuring Automatic File Creation and Deletion</title>
74 <para>There are several services that create or delete files or
78 <listitem><para>systemd-tmpfiles-clean.service</para></listitem>
79 <listitem><para>systemd-tmpfiles-setup-dev.service</para></listitem>
80 <listitem><para>systemd-tmpfiles-setup.service</para></listitem>
83 <para>The system location for the configuration files is
84 <filename>/usr/lib/tmpfiles.d/*.conf</filename>. The local
85 configuration files are in
86 <filename class="directory">/etc/tmpfiles.d</filename>. Files in
87 <filename class="directory">/etc/tmpfiles.d</filename> override
88 files with the same name in
89 <filename class="directory">/usr/lib/tmpfiles.d</filename>. See
90 <filename>tmpfiles.d(5)</filename> manual page for file format
94 Note that the syntax for the
95 <filename>/usr/lib/tmpfiles.d/*.conf</filename> files can be
96 confusing. For example, the default deletion of files in the /tmp directory
97 is located in <filename>/usr/lib/tmpfiles.d/tmp.conf</filename> with
100 <screen role="nodump">q /tmp 1777 root root 10d</screen>
102 The type field, q, discusses creating a subvolume with quotas which
103 is really only applicable to btrfs filesystems. It references type v
104 which in turn references type d (directory). This then creates the
105 specified directory if is is not present and adjusts the permissions
106 and ownership as specified. Contents of the directory will be
107 subject to time based cleanup if the age argument is specified.
111 If the default parameters are not desired, then the file should
112 be copied to <filename class="directory">/etc/tmpfiles.d</filename>
113 and edited as desired. For example:
115 <screen role="nodump"><userinput>mkdir -p /etc/tempfiles.d
116 cp /usr/lib/tmpfiles.d/tmp.conf /etc/tempfiles.d</userinput></screen>
122 <title>Overriding Default Services Behavior</title>
124 <para>The parameter of a unit can be overriden by creating a directory
125 and a configuration file in <filename
126 class="directory">/etc/systemd/system</filename>. For example:</para>
128 <screen role="nodump"><userinput>mkdir -pv /etc/systemd/system/foobar.service.d
130 cat > /etc/systemd/system/foobar.service.d/foobar.conf << EOF
133 RestartSec=30</literal>
134 EOF</userinput></screen>
136 <para>See <filename>systemd.unit(5)</filename> manual page for more
137 information. After creating the configuration file, run
138 <userinput>systemctl daemon-reload</userinput> and <userinput>systemctl
139 restart foobar</userinput> to activate the changes to a service.</para>
144 <title>Debugging the Boot Sequence</title>
146 <para>Rather than plain shell scripts used in SysVinit or BSD style init
147 systems, systemd uses a unified format for different types of startup
148 files (or units). The command <command>systemctl</command> is used to
149 enable, disable, control state, and obtain status of unit files. Here
150 are some examples of frequently used commands:</para>
154 <para><command>systemctl list-units -t <replaceable><service></replaceable> [--all]</command>:
155 lists loaded unit files of type service.</para>
158 <para><command>systemctl list-units -t <replaceable><target></replaceable> [--all]</command>:
159 lists loaded unit files of type target.</para>
162 <para><command>systemctl show -p Wants <replaceable><multi-user.target></replaceable></command>:
163 shows all units that depend on the multi-user target. Targets are
164 special unit files that are anogalous to runlevels under
168 <para><command>systemctl status <replaceable><servicename.service></replaceable></command>:
169 shows the status of the servicename service. The .service extension
170 can be omitted if there are no other unit files with the same name,
171 such as .socket files (which create a listening socket that provides
172 similar functionality to inetd/xinetd).</para>
179 <title>Working with the Systemd Journal</title>
181 <para>Logging on a system booted with systemd is handled with
182 systemd-journald (by default), rather than a typical unix syslog daemon.
183 You can also add a normal syslog daemon and have both work side by
184 side if desired. The systemd-journald program stores journal entries in a
185 binary format rather than a plain text log file. To assist with
186 parsing the file, the command <command>journalctl</command> is provided.
187 Here are some examples of frequently used commands:</para>
191 <para><command>journalctl -r</command>: shows all contents of the
192 journal in reverse chronological order.</para>
195 <para><command>journalctl -u <replaceable>UNIT</replaceable></command>:
196 shows the journal entries associated with the specified UNIT
200 <para><command>journalctl -b[=ID] -r</command>: shows the journal
201 entries since last successful boot (or for boot ID) in reverse
202 chronological order.</para>
205 <para><command>journalctl -f</command>: povides functionality similar
206 to tail -f (follow).</para>
213 <title>Long Running Processes</title>
215 <para>Beginning with systemd-230, all user processes are killed when a user
216 session is ended, even if nohup is used, or the process uses the
217 <function>daemon()</function> or <function>setsid()</function> functions.
218 This is a deliberate change from a historically permissive environment to a
219 more restrictive one. The new behavior may cause issues if you depend on
220 long running programs (e.g., <command>screen</command> or
221 <command>tmux</command>) to remain active after ending your user session.
222 There are three ways to enable lingering processes to remain after a user
223 session is ended.</para>
228 <emphasis>Enable process lingering for only selected users</emphasis>:
229 Normal users have permission to enable process lingering
230 with the command <command>loginctl enable-linger</command> for their
231 own user. System administrators can use the same command with a
232 <parameter>user</parameter> argument to enable for a user. That user
233 can then use the <command>systemd-run</command> command to start
234 long running processes. For example: <command>systemd-run --scope
235 --user /usr/bin/screen</command>. If you enable lingering for your
236 user, the user@.service will remain even after all login sessions are
237 closed, and will automatically start at system boot. This has the
238 advantage of explicitly allowing and disallowing processes to run
239 after the user session has ended, but breaks backwards compatibility
240 with tools like <command>nohup</command> and utilities that use
241 <function>deamon()</function>.
246 <emphasis>Enable system-wide process lingering</emphasis>:
247 You can set <parameter>KillUserProcesses=no</parameter> in
248 <filename>/etc/logind.conf</filename> to enable process lingering
249 globally for all users. This has the benefit of leaving the old
250 method available to all users at the expense of explicit control.
255 <emphasis>Disable at build-time</emphasis>: You can enable
256 lingering by default while building systemd by adding the switch
257 <parameter>-Ddefault-kill-user-processes=no</parameter> to the
258 <command>meson</command> command for systemd. This completely
259 disables the ability of systemd to kill user processes at session