Introduction to systemd
Basics
- WHAT?
systemd
is used to manage system settings and services.systemd
organizes tasks into components called units and groups of units into targets.- WHY?
Learn about the basics of
systemd
, which include essential functionalities such as service management, dependency tracking, logging, resource management, socket activation and system control.- EFFORT
20 minutes of reading time.
- REQUIREMENTS
Basic understanding of Linux commands
Basic understanding of Linux processes, daemons, and control groups
1 What is systemd
? #
systemd
is a system and service manager for Linux operating systems. It is the default initialization system for major Linux distributions.
systemd
is not directly initiated by the user, but installed through the /sbin/init
and started during the early boot.
systemd
acts as the init system that brings up and maintains user space services when run as the first process on boot (PID 1).
PID 1 is known as init and is the first Linux user-mode process created.
It runs until the system shutdown.
systemd
owns PID 1, and is started directly by the kernel. All other processes are started directly
by systemd
or one of its child processes. systemd
mounts the host's file system and manages temporary files.
It is backward compatible with the SysV init scripts. SysV is an initialization system that predates systemd
.
In systemd
, a unit is a resource that the system knows how to operate on and manage. This is
the primary object that the systemd
tools use. These resources are defined with configuration files called unit files.
systemctl
is the central management tool for controlling the init system. It is used to
examine and control the state of the systemd
system and service manager.
Targets in systemd
are groups of related units that act as synchronization points during a system boot.
Target unit files have a .target
file extension. Target units group together various systemd
units through a chain of dependencies.
For troubleshooting, you can use journalctl
, which is
used to query and display log messages from the systemd
journal.
For more information on systemd
, you can refer to https://systemd.io and man 1 systemd
.
2 About the systemd
boot process #
The first step in the boot process is to load the Linux kernel, which is the main component of the Linux operating system.
Once the kernel is loaded, it initializes the hardware and starts the systemd
process, which is the first
process that runs on the system.
2.1 Linux boot process #
The Linux boot process is the initial stage of the operating system's startup. It is the process by which the operating system loads the memory, initializes the components and prepares to execute user applications.
The Linux boot process is divided into four main stages:
- Stage 1: BIOS
When you power on your computer, your computer starts BIOS (Basic Input/Output System) and performs a POST (Power On Self Test). This is an integrity check that probes the hardware functionality of components such as hard disks, SSD, keyboard, RAM, USB ports and any other hardware. If the hardware works as expected, the boot process moves on to the next stage.
- Stage 2: The boot loader
Once POST is complete, BIOS searches for and loads the boot loader program stored in the MBR (Master Boot Record). The MBR is a 512-byte code that is usually located at
/dev/sda
or/dev/hda
depending on your hard drive architecture. The MBR can also be located on a live USB or DVD installation of Linux. BIOS loads and executes this MBR code.There are three main boot loaders in Linux: LILO, GRUB and GRUB2. The GRUB2 (Grand Unified Bootloader) boot loader is the latest and primary boot loader in modern Linux distributions. The GRUB2 configuration file is located at
/boot/grub2/grub2.cfg
. Once BIOS locates the GRUB2 boot loader, It executes and loads it into the main memory (RAM).- Stage 3: Linux kernel initialization
The Linux kernel is the heart of the operating system. In your Linux system, the kernel interfaces with the hardware, controls memory management and manages processes. The boot loader loads the selected Linux kernel. The kernel self extracts from a compressed version and mounts the root file system. It then runs the
/sbin/init
program.- Stage 4:
systemd
The kernel loads
systemd
, which is a system and service manager for Linux operating systems.systemd
then runs all the other initialization processes.
2.2 Boot process with systemd
#
Once the kernel loads systemd
, systemd
takes over and starts the other system services that are required
to bring the system up and running. This includes services such as networking service, the login manager, etc.
The boot process is parallelized in the order in which specific target units are executed.
systemd
uses the /etc/systemd/system/default.target
file to determine the target that the Linux system should boot into.
This file is a link to graphical.target
which boots the graphical login manager.
systemd
activates all target units that are dependencies of default.target
as well as recursively all dependencies of these dependencies.
Once all the services are started, your system is ready to use and the login manager displays. You can now log in and start using the system.
2.3 Analyzing the system boot process performance with the systemd-analyze
command #
Use the systemd-analyze
command to analyze the performance of the system
boot process. The command can also be used to retrieve other state and tracing information
from the system and service manager. It is used to check that unit files are correct and also to access special functions
useful for advanced system manager debugging.
Some examples include:
- Viewing the time it takes for the system to boot
>
systemd-analyze time Startup finished in 3.404s (kernel) + 2.415s (initrd) + 13.125s (userspace) = 18.945s graphical.target reached after 13.117s in userspace- Getting a high-level overview of the boot process, which includes services that are started and the time it takes for each service to start
>
systemd-analyze critical-chain The time when unit became active or started is printed after the "@" character. The time the unit took to start is printed after the "+" character. graphical.target @13.117s └─multi-user.target @13.117s └─getty.target @13.117s └─getty@tty1.service @13.116s └─plymouth-quit-wait.service @10.775s +2.338s └─systemd-user-sessions.service @10.769s +3ms └─remote-fs.target @10.764s └─iscsi.service @10.747s +16ms └─network-online.target @10.744s └─NetworkManager-wait-online.service @1.547s +9.197s └─NetworkManager.service @1.507s +37ms └─network-pre.target @1.504s └─wpa_supplicant.service @2.341s +5ms └─dbus.service @1.042s └─basic.target @1.036s └─sockets.target @1.036s └─snapd.socket @1.035s +590us └─sysinit.target @1.030s └─systemd-update-utmp.service @1.025s +5ms └─auditd.service @976ms +47ms └─systemd-tmpfiles-setup.service @964ms +9ms └─local-fs.target @962ms └─snapd.mounts.target @961ms └─snap-core18-2796.mount @417ms +543ms └─dev-loop9.device @961ms +628usThis command prints a tree of time-critical units either for each of the specified units or for the default target. The initialization of services might depend on socket activation and the parallel execution of units. Similar to the
blame
command, it displays the time taken by a unit to activate, which is not defined for units like device units that transition directly to active state.- Viewing a list of services started during the boot process and displayed according to the time taken by each service
>
systemd-analyze blame 9.197s NetworkManager-wait-online.service 4.002s fwupd.service 2.338s plymouth-quit-wait.service 1.282s dracut-pre-udev.service 1.062s sys-devices-platform-serial8250-tty-ttyS0.device 1.062s dev-ttyS0.device 1.061s dev-ttyS1.device 1.061s sys-devices-platform-serial8250-tty-ttyS1.device 1.060s dev-ttyS11.device 1.060s sys-devices-platform-serial8250-tty-ttyS11.device 1.059s sys-devices-platform-serial8250-tty-ttyS13.device 1.059s dev-ttyS13.device 1.059s sys-devices-platform-serial8250-tty-ttyS10.device 1.059s dev-ttyS10.device 1.058s sys-devices-platform-serial8250-tty-ttyS14.device 1.058s dev-ttyS14.device 1.058s dev-ttyS12.device 1.058s sys-devices-platform-serial8250-tty-ttyS12.device 1.056s sys-devices-platform-serial8250-tty-ttyS17.deviceThe initialization of one service might be slow because it is waiting for another service initialization to complete. It displays the time taken by a unit to activate, which is not defined for units like device units that transition directly to active state. This command does not display results for services with
Type=simple
becausesystemd
considers these services to be started immediately, hence the initialization delays cannot be analyzed.- Generating a vector graphics file that displays the events that take place during the boot process
>
systemd-analyze plot > /temp/sample.svgThis command creates a SVG file in the
temp
directory. The SVG file is a text file that defines a set of graphics vectors that applications such as LibreOffice Draw use to generate a graph.
3 Structure of a unit file #
In systemd
, a unit refers to any resource that the system knows how to operate on and manage.
This is the primary object that the systemd
tools use. These resources are defined using configuration files
called unit files. Administration is easier when you understand unit files when working with systemd
.
Unit files use a simple declarative syntax that allows you to see easily the purpose and effects of a
unit upon activation. Unit files have sections with directives, for example:
[Section] Directive1=value Directive2=value . . .
Unit file types include the following sections:
[Unit]
The first section found in most unit files is the
[Unit]
section. This section is used to define the metadata of the unit file and configure the relationship of the unit file to other unit files. This section is usually placed at the top because it provides an overview of the unit file.[Automount] / [Mount] / [Path] / [Service] / [Slice] / [Socket] /[Swap] / [Timer]
Sections containing directives that are specific to the respective type. See Section 4, “Unit file types” for a list of available types. Note that the types
device
,target
,snapshot
andscope
do not have a type-specific section.[Install]
This is often the last section in the unit file and is optional. This section is used to define the behavior of a unit file when it is enabled or disabled. When you enable a unit file, it automatically starts at boot. Based on the specific unit, there could be a dependency on other related units to work properly. For example,
chrony
requires the directivesAfter
,Wants
, andBefore
, which are all dependencies forchrony
to work with.
systemd
service file #[Unit] Description=usbguard 1 [Service] ExecStart=/usr/sbin/usb-daemon 2 [Install] WantedBy=multi-user.target 3
4 Unit file types #
You can determine the type of unit by its file extension. systemd
categorizes units according to the type of resource they describe.
Types of unit files available for systemd
:
.service
Describes how to manage a service or application. This includes how to start or stop the service, reload its configuration file (if applicable), under what conditions the service starts automatically, and the dependency or the hierarchy information for related unit files.
.scope
This unit file is automatically created by
systemd
from the information received from the D-Bus interface and is used to manage sets of system processes that are created externally..path
Defines a path for path-based activation. By default, a
.service
unit file of the same base name is activated.inotify
is a kernel API that is used by programs that want to be notified about changes to files..snapshot
The
systemctl snapshot
command automatically creates a.snapshot
unit file. This command creates temporary snapshots of the current state of the system. You can modify the current state of the system after making changes. Snapshots are used for rolling back temporary states..timer
Defines a timer that is managed by
systemd
. This is similar to a cron job for delayed or scheduled activation. A unit file with the same name, but with file extension.service
is started when the timer is reached..slice
Associate Linux Control Group nodes, which allow resources to be assigned or restricted to any processes associated with the slice. The name indicates the hierarchy within the control group tree. Units are placed in slices by default depending on their type.
.target
Provides synchronization for other units during a boot up or a change in state, or brings the system to a new state. Other units specify their relation to targets in order to sync with the target's operations.
.socket
Describes a network, an IPC socket, or a FIFO buffer that
systemd
uses for socket-based activation. There is an associated.service
file that starts when an activity is seen on the socket that this unit defines..device
Defines a device that has been designated for
systemd
management byudev
orsysfs
file system. Not all devices have the.device
file. This unit file is required when ordering, mounting, or accessing a device..swap
Defines the swap space on the system. The name of the unit file must reflect the device or file path of the space.
.mount
Defines a mount point on the system to be managed by
systemd
. This file is named after the mount path, with the slashes changed to dashes. Entries within/etc/fstab
can have units created automatically..automount
Defines a mount point that is automatically mounted. Name the file after the mount point that it refers to. A matching
.mount
unit file is required to define the specifics of the mount.
5 Unit dependencies and order #
systemd
has two types of dependencies: requirement and order dependencies. Requirement dependencies specify
which other units must be either started or stopped when activating a unit.
Order dependencies specify the order in which units must be started.
Unit dependencies
Unit files have the dependencies feature. A unit may want or require one or more other units before it can run.
These dependencies are set in unit files with the directives Wants
and Requires
.
Wants
For example, if unit A has
Wants=unit B
, when unit A is run, unit B runs as well. But if unit B starts successfully or not, does not have an influence on unit A running successfully.Requires
If unit A has
Requires=unit B
, both units run but if unit B does not run successfully, unit A is deactivated. It does not matter if the processes of unit A would have run successfully.
Unit order
Without proper instructions, systemd
can run a group of units at the same time. Starting services in the right order
is important for a good functioning of the Linux system. You can arrange the order with the unit file directives Before
and After
.
Before
For example, if unit A has
Before=unit B
, when both units are run, unit A is executed fully before unit B.After
If unit A has
After=unit B
, when both units are run, unit B is executed fully before unit A.
6 Logging #
Log files and journals are important for system administration. They give in-depth information about a system and are very important
for troubleshooting and auditing. Log files contain events and messages generated by the kernel, applications, and users that log into the system.
You can use the journalctl
command to query the journal. This command views logs collected by systemd
.
The systemd-journald
service handles systemd
's log collection. systemd-journald
saves the events and messages in a binary format.
7 systemd
targets #
systemd
uses units and targets. A systemd
unit defines a service or action on the system, which consists of a name, type, and configuration file.
A systemd
target combines several units and defines which services have to be started to reach the target.
On a server, for example, this is a state where the network is running and multiple users can log in.
These files are identified by the suffix .target
.
Similar to unit files, different targets may be nested via dependencies. For example, multi-user.target
requires (among others) the targets
that set up login and user session services.
Common systemd
targets:
-
default.target
Boots by default. The
default.target
file is a symbolic link to the true target file, such asgraphical.target
for a desktop workstation. For a server, it is usuallygraphical.target
.-
poweroff.target
Shuts down and powers off the system.
-
rescue.target
Target unit that pulls the base system and starts a rescue shell session.
-
multi-user.target
Sets up a non-graphical (console) multi-user system.
-
graphical.target
Uses a graphical multi-user system with network services.
-
reboot.target
Shuts down and reboots the system.
For more information on systemd
targets, refer to man 5 systemd.target
and man 7 systemd.special.
8 Using systemd
as a regular user #
You can use systemd
as a regular user for better security or when you do not have root
user privileges. Running an unprivileged service can be done by creating a user
service.
When creating and using a user service, consider the following:
User service sessions are terminated when the user's session ends. This can be overridden by using the
loginctl enable-linger USERNAME
command.User service files are located in
/etc/systemd/user
or$HOME/.config/systemd/user/
.You can control user services with the
systemctl --user
command.
9 systemctl
commands overview #
The systemctl
command is used to examine and control
the state of systemd
and service manager.
You can use the following common systemctl
commands and
refer to the man systemctl page.
9.1 Viewing systemd
information #
To view information about systemd
components, you can use the following
commands:
- systemctl list-units
Lists the
systemd
units. You can use the optional arguments:--state=running
to show the active units and--type=service
to show the exited and active units.- systemctl list-unit-files
Lists the
systemd
units and the status, such as static, generated, disabled, alias, masked, and enabled.- systemctl list-dependencies
Lists the dependency tree.
- systemctl list-dependencies UNIT_FILE
Lists the dependencies of a unit file.
9.2 Managing systemd
services #
The systemctl
command enables you to perform the
following tasks with services.
- systemctl status SERVICE
Checks the status of the specific service.
- systemctl show SERVICE
Displays the service information.
- systemctl start SERVICE
Instead of manually starting the service, use the
start
command. When a change is made to the configuration file, the related service must be started again.- systemctl stop SERVICE
Stops a specific running service.
- systemctl restart SERVICE
Instead of manually restarting the service, use the
restart
command. When a change is made to the configuration file, the related service must be restarted again.- systemctl enable SERVICE
Enables the service on boot.
- systemctl disable SERVICE
Disables the service on boot.
- systemctl reload-or-restart SERVICE
Reload the service if it supports reloading, otherwise it restarts the service. If the service is not running, it is restarted.
- systemctl mask SERVICE
When a service is masked, this means the unit file is symlinked to
/dev/null
. A symlink for a masked service is created from/etc/systemd/system
to point to/dev/null
. This makes it impossible to load the service even if another enabled service requires it. It must be stopped manually, or it continues to run in the background. You can use--runtime
option to only mask temporarily until the next reboot of the system.Created symlink /etc/systemd/system/FOSSLinux.service → /dev/null.
- systemctl unmask SERVICE
Unmasks the service. It is effective when the system is started or restarted manually.
9.3 Managing system states #
The systemctl
command enables you to perform power
management processes on your system, like restarting, shutting down and
so on, as described below.
- systemctl reboot
Reboots the system
reboot.target
.- systemctl poweroff
Powers off the system
poweroff.target
.- systemctl emergency
Goes into the emergency mode
emergency.target
.- systemctl default
Goes back to default target
multi-user.target
.- systemctl soft-reboot
The command reboots only the user space. It shuts down all services and units, restarts
systemd
and starts allsystemd
services again. It can also switch to a new root file system. On the contrary, hardware, kernel space and initrd are not restarted.
10 systemd
troubleshooting #
You can use the following troubleshooting tips to identify and resolve issues
with systemd
services and ensure a smooth system operation.
- Check the syntax of your
systemd
unit file with thesystemd-analyze verify SERVICE
Before starting or enabling a
systemd
service, check the syntax of the unit file to ensure there are no errors. For example:>
sudo
systemd-analyze verify /etc/systemd/system/my-custom-service.serviceThe command analyzes the unit file and reports any syntax errors, missing files, or other issues. You must fix any reported issues before enabling and starting the service.
- Check the logs for your service with the
journalctl -u SERVICE
command If you experience any issue with a
systemd
service, check the service's log. For example:>
sudo
journalctl -u my-custom-service.serviceThe command displays logs for the specified service, including any error messages, warnings, or other relevant information. You can use these logs to identify and fix issues with the service.
- Use the
systemd-analyze plot
command to visualize the boot process If a service is causing issues during the boot process, you can use the
systemd-analyze plot command
to visualize the boot process and identify issues. For example:>
sudo
systemd-analyze plot > boot-plot.svgThe command creates an SVG file called
boot-plot.svg
that contains a graphical representation of the boot process and potential issues. This includes the start and stop time of each service. You can open this file in an SVG-compatible image viewer or Web browser to analyze services that are causing issues during the boot process.- Troubleshoot failed services
To find out which services have failed and to inspect the log output:
>
sudo
systemctl --state=failed- Check the runtime status of a service
To find out the current runtime status of a service:
>
sudo
systemctl status SERVICE- Shutdown or reboot takes long
If the shutdown or reboot takes long, it could be a service that is not exiting.
systemd
waits for some time for each service to exit before trying to terminate it. A common issue is a suspended service or a stalled shutdown. To find out, use the following:>
sudo
systemctl poweroff Failed to power off system via logind: There's already a shutdown or sleep operation in progress>
sudo
systemctl list-jobsYou can cancel the running and waiting jobs, and again shut down or reboot:
>
sudo
systemctl cancel>
sudo
systemctl stop systemd-suspend.service
11 systemd
best practices #
You can follow some of the best practices to ensure efficient systemd
services that are equipped to handle different situations.
- Check the runtime status of a service
To find out the current runtime status of a service:
>
sudo
systemctl status SERVICE- Use absolute path in your
systemd
unit file Use an absolute path for executable files and required files, such as configuration files or scripts in your
systemd
unit file.systemd
does not rely on the user's environment variables like$PATH
to locate files.- Use the ExecReload directive
Use the ExecReload directive in the
[SERVICE]
section when you want to define a specific command that should be executed when you reload a service with thesystemctl reload
command. This is useful for services that can dynamically reload their configuration without a restart.[Service] ExecStart=PATH_TO_EXECUTABLE ExecReload=PATH_TO_RELOAD_SCRIPT
- Use the RestartSec directive
Use the RestartSec directive in the
[SERVICE]
section when you want to define a delay (in seconds) before the service is restarted after a failure. This is useful for services that require a specified time to release resources or prevent rapid restart loops that can cause high system load.[Service] ExecStart=PATH_TO_EXECUTABLE Restart=on-failure RestartSec=5
- Disable emergency mode on a remote machine
You can disable emergency mode on a remote machine, for example, a virtual machine hosted on Google Cloud. If this mode is enabled, the machine is blocked from connecting to the network. For example:
>
sudo
systemctl mask emergency.service>
sudo
systemctl mask emergency.target
12 Legal Notice #
Copyright© 2006–2024 SUSE LLC and contributors. All rights reserved.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or (at your option) version 1.3; with the Invariant Section being this copyright notice and license. A copy of the license version 1.2 is included in the section entitled “GNU Free Documentation License”.
For SUSE trademarks, see https://www.suse.com/company/legal/. All other third-party trademarks are the property of their respective owners. Trademark symbols (®, ™ etc.) denote trademarks of SUSE and its affiliates. Asterisks (*) denote third-party trademarks.
All information found in this book has been compiled with utmost attention to detail. However, this does not guarantee complete accuracy. Neither SUSE LLC, its affiliates, the authors, nor the translators shall be held liable for possible errors or the consequences thereof.