Chapter 3. The system initialization

Table of Contents

3.1. An overview of the boot strap process
3.1.1. Stage 1: the BIOS
3.1.2. Stage 2: the boot loader
3.1.3. Stage 3: the mini-Debian system
3.1.4. Stage 4: the normal Debian system
3.2. Systemd init
3.2.1. The hostname
3.2.2. The filesystem
3.2.3. Network interface initialization
3.2.4. The kernel message
3.2.5. The system message
3.2.6. System management under systemd
3.2.7. Customizing systemd
3.3. The udev system
3.3.1. The kernel module initialization

It is wise for you as the system administrator to know roughly how the Debian system is started and configured. Although the exact details are in the source files of the packages installed and their documentations, it is a bit overwhelming for most of us.

I did my best to provide a quick overview of the key points of the Debian system and their configuration for your reference, based on the current and previous knowledge of mine and others. Since the Debian system is a moving target, the situation over the system may have been changed. Before making any changes to the system, you should refer to the latest documentation for each package.

[Tip] Tip

bootup(7) describes the system bootup process based on systemd . (Recent Debian)

[Tip] Tip

boot(7) describes the system bootup process based on UNIX System V Release 4. (Older Debian)

The computer system undergoes several phases of boot strap processes from the power-on event until it offers the fully functional operating system (OS) to the user.

For simplicity, I limit discussion to the typical PC platform with the default installation.

The typical boot strap process is like a four-stage rocket. Each stage rocket hands over the system control to the next stage one.

Of course, these can be configured differently. For example, if you compiled your own kernel, you may be skipping the step with the mini-Debian system. So please do not assume this is the case for your system until you check it yourself.

[Note] Note

For non-legacy PC platform such as the SUN or the Macintosh system, the BIOS on ROM and the partition on the disk may be quite different (Section 9.5.2, “Disk partition configuration”). Please seek the platform specific documentations elsewhere for such a case.

The BIOS is the 1st stage of the boot process which is started by the power-on event. The BIOS residing on the read only memory (ROM) is executed from the particular memory address to which the program counter of CPU is initialized by the power-on event.

This BIOS performs the basic initialization of the hardware (POST: power on self test) and hands the system control to the next step which you provide. The BIOS is usually provided with the hardware.

The BIOS startup screen usually indicates what key(s) to press to enter the BIOS setup screen to configure the BIOS behavior. Popular keys used are F1, F2, F10, Esc, Ins, and Del. If your BIOS startup screen is hidden by a nice graphics screen, you may press some keys such as Esc to disable this. These keys are highly dependent on the hardware.

The hardware location and the priority of the code started by the BIOS can be selected from the BIOS setup screen. Typically, the first few sectors of the first found selected device (hard disk, floppy disk, CD-ROM, …) are loaded to the memory and this initial code is executed. This initial code can be any one of the following.

  • The boot loader code

  • The kernel code of the stepping stone OS such as FreeDOS

  • The kernel code of the target OS if it fits in this small space

Typically, the system is booted from the specified partition of the primary hard disk partition. First 2 sectors of the hard disk on legacy PC contain the master boot record (MBR). The disk partition information including the boot selection is recorded at the end of this MBR. The first boot loader code executed from the BIOS occupies the rest of this MBR.

The boot loader is the 2nd stage of the boot process which is started by the BIOS. It loads the system kernel image and the initrd image to the memory and hands control over to them. This initrd image is the root filesystem image and its support depends on the bootloader used.

The Debian system normally uses the Linux kernel as the default system kernel. The initrd image for the current 2.6/3.x Linux kernel is technically the initramfs (initial RAM filesystem) image. The initramfs image is a gzipped cpio archive of files in the root filesystem.

[Warning] Warning

The above is no longer true with the new multi-segment initramfs. See Bug #790100.

The default install of the Debian system places first-stage GRUB boot loader code into the MBR for the PC platform. There are many boot loaders and configuration options available.


[Warning] Warning

Do not play with boot loaders without having bootable rescue media (USB memory stick, CD or floppy) created from images in the grub-rescue-pc package. It makes you boot your system even without functioning bootloader on the hard disk.

For GRUB Legacy, the menu configuration file is located at "/boot/grub/menu.lst". For example, it has entries as the following.

title           Debian GNU/Linux
root            (hd0,2)
kernel          /vmlinuz root=/dev/hda3 ro
initrd          /initrd.img

For GRUB 2, the menu configuration file is located at "/boot/grub/grub.cfg". It is automatically generated by "/usr/sbin/update-grub" using templates from "/etc/grub.d/*" and settings from "/etc/default/grub". For example, it has entries as the following.

menuentry "Debian GNU/Linux" {
        set root=(hd0,3)
        linux /vmlinuz root=/dev/hda3
        initrd /initrd.img
}

For these examples, these GRUB parameters mean the following.


[Note] Note

The value of the partition number used by GRUB legacy program is one less than normal one used by Linux kernel and utility tools. GRUB 2 program fixes this problem.

[Tip] Tip

UUID (see Section 9.5.3, “Accessing partition using UUID”) may be used to identify a block special device instead of its file name such as "/dev/hda3", e.g."root=UUID=81b289d5-4341-4003-9602-e254a17ac232 ro".

[Tip] Tip

If GRUB is used, the kernel boot parameter is set in /boot/grub/grub.cfg. On Debian system, you should not edit /boot/grub/grub.cfg directly. You should edit the GRUB_CMDLINE_LINUX_DEFAULT value in /etc/default/grub and run update-grub(8) to update /boot/grub/grub.cfg.

[Tip] Tip

You can start a boot loader from another boot loader using techniques called chain loading.

See "info grub" and grub-install(8).

The mini-Debian system is the 3rd stage of the boot process which is started by the boot loader. It runs the system kernel with its root filesystem on the memory. This is an optional preparatory stage of the boot process.

[Note] Note

The term "the mini-Debian system" is coined by the author to describe this 3rd stage boot process for this document. This system is commonly referred as the initrd or initramfs system. Similar system on the memory is used by the Debian Installer.

The "/init" program is executed as the first program in this root filesystem on the memory. It is a program which initializes the kernel in user space and hands control over to the next stage. This mini-Debian system offers flexibility to the boot process such as adding kernel modules before the main boot process or mounting the root filesystem as an encrypted one.

  • The "/init" program is a shell script program if initramfs was created by initramfs-tools. You can interrupt this part of the boot process to gain root shell by providing "break=init" etc. to the kernel boot parameter. See the "/init" script for more break conditions. This shell environment is sophisticated enough to make a good inspection of your machine's hardware. Commands available in this mini-Debian system are stripped down ones and mainly provided by a GNU tool called busybox(1).

  • The "/init" program is a binary systemd program if initramfs was created by dracut. ** Commands available in this mini-Debian system are stripped down systemd(1) environment.

[Caution] Caution

You need to use "-n" option for mount command when you are on the readonly root filesystem.

The normal Debian system is the 4th stage of the boot process which is started by the mini-Debian system. The system kernel for the mini-Debian system continues to run in this environment. The root filesystem is switched from the one on the memory to the one on the real hard disk filesystem.

The init program is executed as the first program with PID=1 to perform the main boot process of starting many programs. The default file path for the init program is "/sbin/init" but it can be changed by the kernel boot parameter as "init=/path/to/init_program".

The default init program has been changing:

  • Debian before squeeze uses the simple SysV-style init.

  • Debian wheezy improves the SysV-style init by ordering the boot sequence with LSB header and starting boot scripts in parallel.

  • Debian jessie switches its default init to the systemd for the event-driven and parallel initialization.

[Tip] Tip

The actual init command on your system can be verified by the "ps --pid 1 -f" command.

[Tip] Tip

"/sbin/init" is symlinked to "/lib/systemd/systemd" after Debian jessie.


[Tip] Tip

See Debian wiki: BootProcessSpeedup for the latest tips to speed up the boot process.

This section describes how system is started by the systemd(1) program with PID=1 (i.e., init process).

The systemd init process spawns processes in parallel based on the unit configuration files (see systemd.unit(5)) which are written in declarative style instead of SysV-like procedural style. These are loaded from a set of paths (see systemd-system.conf(5)) as follows:

  • "/lib/systemd/system": OS default configuration files

  • "/etc/systemd/system": system administrator configuration files which override the OS default configuration files

  • "/run/systemd/system": run-time generated configuration files which override the installed configuration files

Their inter-dependencies are specified by the directives "Wants=", "Requires=", "Before=", "After=", … (see "MAPPING OF UNIT PROPERTIES TO THEIR INVERSES" in systemd.unit(5)). The resource controls are also defined (see systemd.resource-control(5)).

The suffix of the unit configuration file encodes their types as:

  • *.service describes the process controlled and supervised by systemd. See systemd.service(5).

  • *.device describes the device exposed in the sysfs(5) as udev(7) device tree. See systemd.device(5).

  • *.mount describes the file system mount point controlled and supervised by systemd. See systemd.mount(5).

  • *.automount describes the file system auto mount point controlled and supervised by systemd. See systemd.automount(5).

  • *.swap describes the swap device or file controlled and supervised by systemd. See systemd.swap(5).

  • *.path describes the path monitored by systemd for path-based activation. See systemd.path(5).

  • *.socket describes the socket controlled and supervised by systemd for socket-based activation. See systemd.socket(5).

  • *.timer describes the timer controlled and supervised by systemd for timer-based activation. See systemd.timer(5).

  • *.slice manages resources with the cgroups(7). See systemd.slice(5).

  • *.scope is created programmatically using the bus interfaces of systemd to manages a set of system processes. See systemd.scope(5).

  • *.target groups other unit configuration files to create the synchronization point during start-up. See systemd.target(5).

Upon system start up (i.e., init), the systemd process tries to start the "/lib/systemd/system/default.target (normally symlinked to "graphical.target"). First, some special target units (see systemd.special(7)) such as "local-fs.target", "swap.target" and "cryptsetup.target" are pulled in to mount the filesystems. Then, other target units are also pulled in by the target unit dependencies. For details, read bootup(7).

systemd offers backward compatibility features. SysV-style boot scripts in "/etc/init.d/rc[0123456S].d/[KS]<name>" are still parsed and telinit(8) is translated into systemd unit activation requests.

[Caution] Caution

Emulated runlevel 2 to 4 are all symlinked to the same "multi-user.target".

The mount options of normal disk and network filesystems are set in "/etc/fstab". See fstab(5) and Section 9.5.7, “Optimization of filesystem by mount options”.

The configuration of the encrypted filesystem is set in "/etc/crypttab". See crypttab(5)

The configuration of software RAID with mdadm(8) is set in "/etc/mdadm/mdadm.conf". See mdadm.conf(5).

[Warning] Warning

After mounting all the filesystems, temporary files in "/tmp", "/var/lock", and "/var/run" are cleaned for each boot up.

The systemd offers not only init system but also generic system management functionalities such as journal logging, login management, time management, network management. etc..

The systemd(1) is managed by several commands:

  • the systemctl(1) command controls the systemd system and service manager (CLI),

  • the systemsdm(1) command controls the systemd system and service manager (GUI),

  • the journalctl(1) command queries the systemd journal,

  • the loginctl(1) command controls the systemd login manager, and

  • the systemd-analyze(1) analyzes system boot-up performance.

Here are a list of typical systemd management command snippets. For the exact meanings, please read the pertinent manpages.

Table 3.5. List of typical systemd management command snippets

Operation Type Command snippets
GUI for service manager GUI "systemadm" (systemd-ui package)
List all target unit configuration Unit "systemctl list-units --type=target"
List all service unit configuration Unit "systemctl list-units --type=service"
List all unit configuration types Unit "systemctl list-units --type=help"
List all socket units in memory Unit "systemctl list-sockets"
List all timer units in memory Unit "systemctl list-timers"
Start "$unit" Unit "systemctl start $unit"
Stop "$unit" Unit "systemctl stop $unit"
Reload service-specific configuration Unit "systemctl reload $unit"
Stop and start all "$unit" Unit "systemctl restart $unit"
Start "$unit" and stop all others Unit "systemctl isolate $unit"
Switch to "graphical" (GUI system) Unit "systemctl isolate graphical"
Switch to "multi-user" (CLI system) Unit "systemctl isolate multi-user"
Switch to "rescue" (single user CLI system) Unit "systemctl isolate rescue"
Send kill signal to "$unit" Unit "systemctl kill $unit"
Send kill signal to "$unit" Unit "systemctl kill $unit"
Check if "$unit" service is active Unit "systemctl is-active $unit"
Check if "$unit" service is failed Unit "systemctl is-failed $unit"
Check status of "$unit|$PID|device" Unit "systemctl status $unit|$PID|$device"
Show properties of "$unit|$job" Unit "systemctl show $unit|$job"
Reset failed "$unit" Unit "systemctl reset-failed $unit"
List dependency of all unit services Unit "systemctl list-dependencies --all"
List unit files installed on the system Unit file "systemctl list-unit-files"
Enable "$unit" (add symlink) Unit file "systemctl enable $unit"
Disable "$unit" (remove symlink) Unit file "systemctl disable $unit"
Unmask "$unit" (remove symlink to "/dev/null") Unit file "systemctl unmask $unit"
Mask "$unit" (add symlink to "/dev/null") Unit file "systemctl mask $unit"
Get default-target setting Unit file "systemctl get-default"
Set default-target to "graphical" (GUI system) Unit file "systemctl set-default graphical"
Set default-target to "multi-user" (CLI system) Unit file "systemctl set-default multi-user"
Show job environment Environment "systemctl show-environment"
Set job environment "variable" to "value" Environment "systemctl set-environment variable=value"
Unset job environment "variable" Environment "systemctl unset-environment variable"
Reload all unit files and daemons Lifecycle "systemctl daemon-reload"
Shut down the system System "systemctl poweroff"
Shut down and reboot the system System "systemctl reboot"
Suspend the system System "systemctl suspend"
Hibernate the system System "systemctl hibernate"
View job log of "$unit" Journal "journalctl -u $unit"
View job log of "$unit" ("tail -f" style) Journal "journalctl -u $unit -f"
Show time spent for each initialization steps Analyze "systemd-analyze time"
List of all units by the time to initialize Analyze "systemd-analyze blame"
Load and detect errors in "$unit" file Analyze "systemd-analyze verify $unit"
Track boot process by the cgroups(7) Cgroup "systemd-cgls"
Track boot process by the cgroups(7) Cgroup "ps xawf -eo pid,user,cgroup,args"
Track boot process by the cgroups(7) Cgroup Read sysfs under "/sys/fs/cgroup/systemd/"

Here, "$unit" in the above examples may be a single unit name (suffix such as .service and .target are optional) or, in many cases, multiple unit specifications (shell-style globs "*", "?", "[]" using fnmatch(3) which will be matched against the primary names of all units currently in memory).

System state changing commands in the above examples are typically preceded by the "sudo" to attain the required administrative privilege.

The output of the "systemctl status $unit|$PID|$device" uses color of the dot ("●") to summarize the unit state at a glance.

  • White "●" indicates an "inactive" or "deactivating" state.

  • Red "●" indicates a "failed" or "error" state.

  • Green "●" indicates an "active", "reloading" or "activating" state.

With default installation, many network services (see Chapter 6, Network applications) are started as daemon processes after network.target at boot time by systemd. The "sshd" is no exception. Let's change this to on-demand start of "sshd" as a customization example.

First, disable system installed service unit.

 $ sudo systemctl stop sshd.service
 $ sudo systemctl mask sshd.service

The on-demand socket activation system of the classic Unix services was through the indetd superserver. Under systemd, the equivalent can be enabled by adding *.socket and *.service unit configuration files.

sshd.socket for specifying a socket to listen on

[Unit]
Description=SSH Socket for Per-Connection Servers

[Socket]
ListenStream=22
Accept=yes

[Install]
WantedBy=sockets.target

sshd@.service as the matching service file of sshd.socket

[Unit]
Description=SSH Per-Connection Server

[Service]
ExecStart=-/usr/sbin/sshd -i
StandardInput=socket

Then reload.

 $ sudo systemctl daemon-reload

For Linux kernel 2.6 and newer, the udev system provides mechanism for the automatic hardware discovery and initialization (see udev(7)). Upon discovery of each device by the kernel, the udev system starts a user process which uses information from the sysfs filesystem (see Section 1.2.12, “procfs and sysfs”), loads required kernel modules supporting it using the modprobe(8) program (see Section 3.3.1, “The kernel module initialization”), and creates corresponding device nodes.

[Tip] Tip

If "/lib/modules/<kernel-version>/modules.dep" was not generated properly by depmod(8) for some reason, modules may not be loaded as expected by the udev system. Execute "depmod -a" to fix it.

The name of device nodes can be configured by udev rule files in "/etc/udev/rules.d/". Current default rules tend to create dynamically generated names resulting non-static device names except for cd and network devices. By adding your custom rules similar to what cd and network devices do, you can generate static device names for other devices such as USB memory sticks, too. See "Writing udev rules" or "/usr/share/doc/udev/writing_udev_rules/index.html".

Since the udev system is somewhat a moving target, I leave details to other documentations and describe the minimum information here.

[Tip] Tip

For mounting rules in "/etc/fstab", device nodes do not need to be static ones. You can use UUID to mount devices instead of device names such as "/dev/sda". See Section 9.5.3, “Accessing partition using UUID”.

The modprobe(8) program enables us to configure running Linux kernel from user process by adding and removing kernel modules. The udev system (see Section 3.3, “The udev system”) automates its invocation to help the kernel module initialization.

There are non-hardware modules and special hardware driver modules as the following which need to be pre-loaded by listing them in the "/etc/modules" file (see modules(5)).

The configuration files for the modprobe(8) program are located under the "/etc/modprobes.d/" directory as explained in modprobe.conf(5). (If you want to avoid some kernel modules to be auto-loaded, consider to blacklist them in the "/etc/modprobes.d/blacklist" file.)

The "/lib/modules/<version>/modules.dep" file generated by the depmod(8) program describes module dependencies used by the modprobe(8) program.

[Note] Note

If you experience module loading issues with boot time module loading or with modprobe(8), "depmod -a" may resolve these issues by reconstructing "modules.dep".

The modinfo(8) program shows information about a Linux kernel module.

The lsmod(8) program nicely formats the contents of the "/proc/modules", showing what kernel modules are currently loaded.

[Tip] Tip

You can identify exact hardware on your system. See Section 9.4.3, “Hardware identification”.

[Tip] Tip

You may configure hardware at boot time to activate expected hardware features. See Section 9.4.4, “Hardware configuration”.

[Tip] Tip

You can probably add support for your special device by recompiling the kernel. See Section 9.9, “The kernel”.