Linux Boot Process Step by Step: BIOS/UEFI, GRUB, Kernel, initramfs and systemd

By Ravindra Bagale, trainer LinkedIn Last updated: 25 October 2026

When you switch on a laptop, a lot happens before you see the login screen. Firmware wakes the hardware. A boot loader picks a kernel. The kernel mounts the disk. systemd starts your services. If you know these steps, you can fix a server that will not boot.

What you'll learn in this chapter

The 6 boot stages in order
BIOS vs UEFI, MBR vs GPT
What GRUB does and how to edit it safely
What initramfs is and when to rebuild it
systemd targets, default target and rescue mode
How to read boot logs and find slow boot

1. An everyday example: opening a shop in the morning

Why do we need this?

Sanjay opens his medical shop in Nashik at 8 am. The watchman unlocks the main shutter. Sanjay reads the to-do list. His helper brings the cash box and keys. Then the lights, billing PC and fridge go on, one by one. Only then do customers come in.

What is it?

Watchman = firmware (BIOS/UEFI). To-do list = GRUB. Helper with keys = initramfs. Sanjay = kernel. Lights and billing PC = systemd services. Customers = users logging in.

How to do it (step by step)

Each stage hands over to the next one.

2. Stage 1: BIOS or UEFI firmware

Why do we need this?

The CPU knows nothing when power comes on. Some small program on the board must test the hardware and find a disk to boot from.

What is it?

BIOS is the old firmware. It runs POST (power-on self test) and reads the first 512 bytes of the disk (MBR). UEFI is the new firmware. It reads a small FAT partition called the EFI System Partition (ESP) and runs a .efi file. UEFI supports GPT disks over 2 TB and Secure Boot.

Point	BIOS + MBR	UEFI + GPT
Max disk	2 TB	Huge (8 ZB)
Partitions	4 primary	128
Boot code	First 512 bytes	EFI file on ESP
Secure Boot	No	Yes

How to do it (step by step)

Check if the machine booted with UEFI.

[ -d /sys/firmware/efi ] && echo UEFI || echo BIOS

Output:

UEFI

What it means: The folder exists only on UEFI systems.

See the UEFI boot entries.

sudo efibootmgr

Output:

BootCurrent: 0001
BootOrder: 0001,0000
Boot0000* UEFI PXE
Boot0001* ubuntu

What it means: The firmware will try ubuntu first, then network boot.

Check the partition table type.

sudo fdisk -l /dev/sda | grep 'Disklabel'

Output:

Disklabel type: gpt

What it means: The disk uses GPT.

3. Stage 2: GRUB, the boot loader

Why do we need this?

A server can have many kernels: the new one and an old safe one. Someone must pick one and pass options to it.

What is it?

GRUB2 shows the boot menu, loads the kernel file (vmlinuz) and the initramfs into memory, and passes the kernel command line (like root=UUID=... ro quiet). Its config is /boot/grub/grub.cfg (or /boot/grub2/grub.cfg). Never edit that file by hand. Edit /etc/default/grub and rebuild.

How to do it (step by step)

See the kernel files GRUB can load.

ls /boot

Output:

config-5.15.0-91-generic
grub
initrd.img-5.15.0-91-generic
vmlinuz-5.15.0-91-generic

What it means: vmlinuz is the kernel. initrd.img is the initramfs.

See the command line the kernel got.

cat /proc/cmdline

Output:

BOOT_IMAGE=/boot/vmlinuz-5.15.0-91-generic root=UUID=3e6b... ro quiet splash

What it means: root= tells the kernel which disk is /. ro = mount read-only first.

Change the menu wait time.

sudo sed -i 's/^GRUB_TIMEOUT=.*/GRUB_TIMEOUT=5/' /etc/default/grub

What it means: The setting is saved but not active yet.

Rebuild GRUB config (Ubuntu).

sudo update-grub

Output:

Generating grub configuration file ...
Found linux image: /boot/vmlinuz-5.15.0-91-generic
done

What it means: grub.cfg is regenerated with your change.

Rebuild on RHEL/Amazon Linux.

sudo grub2-mkconfig -o /boot/grub2/grub.cfg

Output:

Generating grub configuration file ...
done

What it means: Same job, different command name.

List kernels and set the default (RHEL).

sudo grubby --info=ALL | grep ^kernel; sudo grubby --set-default /boot/vmlinuz-5.14.0-362.el9.x86_64

Output:

kernel="/boot/vmlinuz-5.14.0-427.el9.x86_64"
kernel="/boot/vmlinuz-5.14.0-362.el9.x86_64"

What it means: Next boot uses the older, known-good kernel.

4. Stage 3: the kernel starts

Why do we need this?

The kernel must find the CPU, memory and devices before any program can run.

What is it?

GRUB jumps into the kernel. The kernel unpacks itself, sets up memory, detects hardware, starts drivers and then runs /init from the initramfs. Its messages go to the kernel ring buffer (dmesg).

How to do it (step by step)

Read the first kernel boot lines.

sudo dmesg | head -2

Output:

[    0.000000] Linux version 5.15.0-91-generic
[    0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-5.15.0-91-generic root=UUID=3e6b... ro

What it means: Time 0.000000 is the moment the kernel started.

Find disk detection lines.

sudo dmesg | grep -i 'nvme0n1'

Output:

[    1.204] nvme nvme0: 2/0/0 default/read/poll queues
[    1.210]  nvme0n1: p1 p14 p15

What it means: The kernel found an NVMe disk with 3 partitions.

See only errors and warnings.

sudo dmesg --level=err,warn

Output:

[    2.511] ACPI Warning: ...

What it means: Short list of hardware problems.

5. Stage 4: initramfs

Why do we need this?

The real root disk may need a driver (NVMe, RAID, LVM, encryption) that is stored on that same disk. Chicken and egg. initramfs solves it.

What is it?

initramfs is a small compressed file system loaded into RAM by GRUB. It holds just enough drivers and tools to find and mount the real root file system. Then it switches root (switch_root) and starts systemd. Rebuild it with update-initramfs (Debian) or dracut (RHEL).

How to do it (step by step)

See what is inside the initramfs.

lsinitramfs /boot/initrd.img-$(uname -r) | grep nvme

Output:

usr/lib/modules/5.15.0-91-generic/kernel/drivers/nvme/host/nvme.ko

What it means: The NVMe driver is packed inside, so the root disk can be found.

Rebuild initramfs on Ubuntu.

sudo update-initramfs -u

Output:

update-initramfs: Generating /boot/initrd.img-5.15.0-91-generic

What it means: Needed after adding a disk driver or changing /etc/crypttab.

Rebuild on RHEL/Amazon Linux.

sudo dracut -f

What it means: -f overwrites the current initramfs image.

6. Stage 5: systemd and targets

Why do we need this?

A server has dozens of services. Some need the network, some need the disk. Starting them in the right order, in parallel, is hard by hand.

What is it?

systemd is PID 1. It reads unit files and starts services in parallel based on dependencies. A target is a group of units, like a goal. multi-user.target = server with network, no GUI. graphical.target = with desktop. rescue.target = single user for repair. emergency.target = bare minimum.

Target	Old runlevel	Use
poweroff.target	0	Shut down
rescue.target	1	Repair, single user
multi-user.target	3	Servers
graphical.target	5	Desktops
reboot.target	6	Restart

How to do it (step by step)

See the default target.

systemctl get-default

Output:

multi-user.target

What it means: This server boots to text mode with network.

Set the default target.

sudo systemctl set-default multi-user.target

Output:

Created symlink /etc/systemd/system/default.target → /lib/systemd/system/multi-user.target.

What it means: The default target is just a symlink.

Switch to rescue mode now (careful).

sudo systemctl isolate rescue.target

What it means: Stops most services. Use only on a console, not over SSH.

List failed units after boot.

systemctl --failed

Output:

  UNIT          LOAD   ACTIVE SUB    DESCRIPTION
● nginx.service loaded failed failed A high performance web server

What it means: nginx failed to start. Check its logs next.

7. Reading boot logs and finding slow boots

Why do we need this?

A boot problem leaves clues in the logs. Without logs you are guessing.

What is it?

journalctl reads systemd's journal. -b = this boot, -b -1 = the previous boot. systemd-analyze shows boot time and which unit was slow.

How to do it (step by step)

Total boot time.

systemd-analyze

Output:

Startup finished in 2.1s (kernel) + 9.8s (userspace) = 11.9s
multi-user.target reached after 9.6s in userspace

What it means: The kernel took 2 s. Services took almost 10 s.

Find the slowest units.

systemd-analyze blame | head -3

Output:

6.012s cloud-init.service
1.230s snapd.service
0.811s systemd-networkd-wait-online.service

What it means: cloud-init is the slowest. Look there first.

Errors from this boot only.

journalctl -b -p err

Output:

Oct 25 09:01:12 web01 nginx[812]: nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address already in use)

What it means: nginx could not start: port 80 already used.

Logs from the previous boot (after a crash).

journalctl -b -1 -n 20

Output:

-- Logs begin ... --
... kernel: Out of memory: Killed process 2211 (java)

What it means: The last boot ended with an out-of-memory kill.

List known boots.

journalctl --list-boots

Output:

-1 8c1d... Fri 2026-10-23 10:00—Sat 2026-10-24 21:14
 0 a73f... Sun 2026-10-25 09:00—Sun 2026-10-25 10:42

What it means: Two boots are stored. 0 is the current one.

8. Real examples from many industries

Why do we need this?

The same Linux skill shows up in every kind of company. Seeing it in many places helps you remember it.

What is it?

Each row is a made-up team. It shows who uses this and how.

How to do it (step by step)

Team and city	How they use it	Command
Anjali, bank ops, Hyderabad	Pins an old kernel with grubby after a bad patch	`grubby --set-default`
Tom, SaaS, Austin	Cuts EC2 boot time by checking cloud-init in blame	`systemd-analyze blame`
Fatima, telecom, Riyadh	Adds a RAID driver and rebuilds initramfs	`dracut -f`
Arjun, retail, Chennai	Reads previous boot logs after a power cut	`journalctl -b -1`
Elena, media, Madrid	Boots kiosks to graphical.target	`systemctl set-default graphical.target`
Kwame, logistics, Accra	Uses the EC2 serial console to fix a server stuck at boot	`aws ec2-instance-connect send-serial-console-ssh-public-key`

9. Common errors and how to fix them

Why do we need this?

Everyone hits these errors. Knowing the cause saves hours.

What is it?

Read the exact message first. Then match it to a row below.

How to do it (step by step)

Error you see	Why it happens	How to fix
`error: no such partition. grub rescue>`	GRUB cannot find /boot (disk changed or partition deleted)	Boot a live/rescue disk, mount root, and reinstall GRUB. `grub-install /dev/sda && update-grub`
`Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0)`	initramfs missing or lacks the disk driver	Pick an older kernel in GRUB, then rebuild initramfs. `update-initramfs -u -k all`
`Gave up waiting for root file system device. ALERT! UUID=... does not exist. Dropping to a shell!`	Wrong root=UUID in GRUB or the disk UUID changed	Compare blkid with /proc/cmdline and fix, then update-grub. `blkid`
`You are in emergency mode. After logging in, type "journalctl -xb"`	A line in /etc/fstab points to a missing disk	Fix or comment that fstab line, or add nofail. `mount -a`
`A start job is running for /dev/disk/by-uuid/... (1min 30s)`	systemd waits for a disk that is not there	Add nofail,x-systemd.device-timeout=10 in fstab. `sudo vi /etc/fstab`
`Failed to start nginx.service`	Service config error or port in use	Read the unit logs. `journalctl -u nginx -b`
`error: you need to load the kernel first.`	GRUB entry has no linux line (bad grub.cfg)	Regenerate grub.cfg, do not hand-edit it. `grub2-mkconfig -o /boot/grub2/grub.cfg`
`/boot: No space left on device (during kernel update)`	Too many old kernels in /boot	Remove old kernels with the package manager. `sudo apt autoremove --purge`
`Verification failed: (0x1A) Security Violation`	Secure Boot blocks an unsigned kernel or module	Use a signed kernel or enrol the key with mokutil. `mokutil --sb-state`

10. Troubleshooting step by step

Why do we need this?

A server that does not boot is scary. Find the stage that failed, then fix only that stage.

What is it?

Go top to bottom. Stop at the first step that shows the problem.

How to do it (step by step)

Get a console: cloud serial console, VM console or physical screen.

Stuck before GRUB menu? It is firmware or disk order.

efibootmgr -v

grub rescue prompt? Reinstall GRUB from a rescue disk.

grub-install /dev/sda

Kernel panic? Choose an older kernel in the GRUB menu.

Dropped to (initramfs) prompt? Check the root UUID.

blkid; cat /proc/cmdline

Emergency mode? Check /etc/fstab.

journalctl -xb | grep -i mount

Boots but a service failed? Check the unit.

systemctl --failed; journalctl -u nginx -b

11. Lab 1: Map the boot of your own server

Why do we need this?

At Kavya Logistics in Indore, the manager asks why servers take 40 seconds to boot. Collect facts for each stage.

What is it?

A full hands-on lab. Run every step on a test Linux machine (an EC2 instance, a VM, or WSL).

How to do it (step by step)

Firmware type.

[ -d /sys/firmware/efi ] && echo UEFI || echo BIOS

Output:

UEFI

Kernel command line.

cat /proc/cmdline

Output:

BOOT_IMAGE=/vmlinuz-6.1.0 root=UUID=7a1c... ro console=ttyS0

initramfs size.

ls -lh /boot/initr*

Output:

-rw------- 1 root root 34M Oct 20 10:02 /boot/initrd.img-6.1.0

Default target.

systemctl get-default

Output:

multi-user.target

Boot time split.

systemd-analyze

Output:

Startup finished in 3.0s (kernel) + 37.2s (userspace) = 40.2s

Find the slow unit.

systemd-analyze blame | head -1

Output:

30.004s systemd-networkd-wait-online.service

See the critical chain.

systemd-analyze critical-chain

Output:

multi-user.target @37.1s
└─network-online.target @37.0s
  └─systemd-networkd-wait-online.service @7.0s +30.0s

Conclusion: waiting for a second network card that has no cable.

networkctl

Output:

IDX LINK TYPE  OPERATIONAL
  2 ens5 ether routable
  3 ens6 ether no-carrier

12. Lab 2: Break and fix fstab safely (test VM only)

Why do we need this?

Practice emergency mode recovery before it happens in production.

What is it?

A full hands-on lab. Run every step on a test Linux machine (an EC2 instance, a VM, or WSL).

How to do it (step by step)

Back up fstab.

sudo cp /etc/fstab /etc/fstab.bak

Add a fake disk line.

echo 'UUID=dead-beef /data ext4 defaults 0 2' | sudo tee -a /etc/fstab

Output:

UUID=dead-beef /data ext4 defaults 0 2

Test without reboot (you will see the error).

sudo mount -a

Output:

mount: /data: can't find UUID=dead-beef.

Fix: add nofail so boot never stops for it.

sudo sed -i 's/defaults 0 2/defaults,nofail 0 2/' /etc/fstab

Or restore the backup.

sudo cp /etc/fstab.bak /etc/fstab && sudo mount -a && echo OK

Output:

OK

Ravindra Bagale's Tip

Before every kernel update or GRUB change, make sure you have console access (serial console on EC2). SSH dies first when boot breaks.

13. Frequently asked questions

Q. What is the difference between initrd and initramfs?

Both are early RAM file systems. initrd was a disk image. initramfs is a cpio archive. People use both names today.

Q. Can I edit grub.cfg directly?

No. It is overwritten on the next update. Edit /etc/default/grub and run update-grub or grub2-mkconfig.

Q. What replaced runlevels?

systemd targets. runlevel 3 = multi-user.target, 5 = graphical.target.

Q. How do I reset a lost root password?

At GRUB, edit the entry, add rd.break (RHEL) or init=/bin/bash, remount / rw, run passwd. On SELinux systems touch /.autorelabel.

Q. Why does EC2 not show a GRUB menu?

The timeout is 0 on cloud images. Use the EC2 serial console to interact.

14. Interview questions and answers

Beginner level

Q. Name the boot stages.

Firmware (BIOS/UEFI) → GRUB → kernel → initramfs → systemd → target/login.

Q. What is GRUB?

The boot loader. It shows the menu, loads the kernel and initramfs and passes kernel options.

Q. How do you see the default target?

systemctl get-default.

Q. What is PID 1?

systemd on most distros.

Mid level level

Q. Why do we need initramfs?

To load drivers (NVMe, LVM, RAID, LUKS) needed to mount the real root before it is available.

Q. BIOS vs UEFI?

BIOS reads MBR, max 2 TB, no Secure Boot. UEFI reads an EFI file from the ESP, supports GPT and Secure Boot.

Q. How do you find what slows down boot?

systemd-analyze, systemd-analyze blame and critical-chain.

Q. How do you read logs from the boot before a crash?

journalctl -b -1 (needs persistent journal in /var/log/journal).

Senior level

Q. A kernel update made 200 servers fail to boot. What is your plan?

Stop the rollout. Use console to boot the previous kernel. Set it default with grubby. Find the cause (missing driver in initramfs). Fix the image, test on canary, then roll forward in waves.

Q. How do you make a server boot even if a data disk is missing?

Use UUIDs, add nofail and x-systemd.device-timeout in fstab, or use systemd mount units with proper dependencies.

Q. Explain Secure Boot in one minute.

UEFI checks signatures from shim to GRUB to kernel to modules. Unsigned code is refused. Custom modules need a MOK key.

Q. How do you recover an EC2 instance that does not boot?

Use serial console. If not possible, stop it, detach the root volume, attach to a helper instance, fix fstab/GRUB, reattach and start.

Try at home

Run systemd-analyze blame and write the 3 slowest services.
Check if your PC or VM uses UEFI.
Run journalctl -b -p warn and read 5 lines.

Practice Tasks

About this scenario

Vidarbha Cotton Mills is a textiles business in Nagpur. Ganesh Wankhede looks after its IT and cloud work. They count money in INR and run their AWS resources in the ap-south-1 Region. In the tasks below you do real jobs for them using what this lesson teaches. All names, account IDs, IP addresses and numbers are made up for practice.

0 / 16

Easy Is the target active?

Why: It tells if the server boots with GUI or not.

What to do: Check whether multi-user.target is active right now on the Vidarbha Cotton Mills server.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Easy Find slow boot units

Why: Fix the slowest first.

What to do: List the units that took the most time during the last boot of the Vidarbha Cotton Mills server, printed straight without the pager.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Easy Critical chain of a unit

Why: Know the baseline.

What to do: Show the critical chain of units that sshd.service waited for during boot.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Easy Kernel version string

Why: root= and other options live here.

What to do: Print the full kernel version string of the Vidarbha Cotton Mills server from the process folder.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Easy List failed units

Why: Quick health check after boot.

What to do: List only failed service units on the Vidarbha Cotton Mills server.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Easy Repair target

Why: Single user mode for fixes.

What to do: Which systemd target replaces old runlevel 1 (single user repair)? Type the full target name.

How:

Write down the numbers the task gives you.
Work it out step by step on paper.
Type only the final answer in the box.
Click Check answer.

Medium Set multi-user target

Why: Servers do not need a GUI.

What to do: Make the Vidarbha Cotton Mills server boot to the graphical desktop by default.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Medium Errors from this boot

Why: Focus only on what matters.

What to do: Show only warning-level (and worse) journal messages from the current boot.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Medium Logs from two boots ago

Why: After a crash, the clue is in the last boot.

What to do: Vidarbha Cotton Mills's server rebooted by itself. Show the journal of the boot before the previous one (two boots ago).

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Medium Rebuild GRUB on Ubuntu

Why: Edits to /etc/default/grub need a rebuild.

What to do: You changed GRUB_TIMEOUT on the Ubuntu server. Regenerate the GRUB config, and only if that works, reboot.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Medium Rebuild initramfs (Ubuntu)

Why: Needed after driver or crypttab changes.

What to do: Update the initramfs for all installed kernels on Ubuntu.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Medium Rebuild initramfs (RHEL)

Why: RHEL uses dracut.

What to do: Force-rebuild the initramfs of every installed kernel on Amazon Linux.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Medium Test fstab without reboot

Why: A bad fstab line drops boot to emergency mode.

What to do: After editing /etc/fstab on the Vidarbha Cotton Mills server, mount everything listed in it to test, printing each mount (verbose).

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Medium Kernel errors only

Why: Hardware problems show here.

What to do: Show kernel ring buffer messages with level crit or alert only.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

Medium Keep booting with a missing disk

Why: Detached EBS volumes are common.

What to do: Which fstab mount option lets the Vidarbha Cotton Mills server finish booting even if a data disk is missing? Type the option.

How:

Write down the numbers the task gives you.
Work it out step by step on paper.
Type only the final answer in the box.
Click Check answer.

Hard Rebuild GRUB on RHEL

Why: Different name on Red Hat.

What to do: On an older RHEL 8 UEFI server, regenerate /boot/efi/EFI/redhat/grub.cfg.

How:

Find the Linux command that does this job.
Add the options (like -r or -h) the task needs.
Add the file names, folders or users in the right order.
Use sudo if the job changes system files or users. Click Check answer.

See all practice tasks in the Task Hub →

In short

Firmware → GRUB → kernel → initramfs → systemd → target.
Edit /etc/default/grub, then update-grub or grub2-mkconfig.
Rebuild initramfs with update-initramfs -u or dracut -f.
journalctl -b and systemd-analyze blame are your boot friends.

Revision, comparison and interview cards

Revise this lesson in a few minutes. Save the cards on your phone or download them as PNG files. The QR code on each card opens this lesson.

Linux boot steps. Know where boot fails. UEFI, GRUB, kernel, initramfs. Then systemd starts services. See slow boot units. — Revision card: Linux boot steps · Download PNG · Open card page

systemd basics. Starts and watches services. PID 1, uses unit files. start, stop, enable, status. Read a service log. — Revision card: systemd basics · Download PNG · Open card page

BIOS vs UEFI. Old firmware. / Modern firmware. — Comparison: BIOS vs UEFI · Download PNG · Open card page

start vs enable. Runs the service now. / Runs it at every boot. — Comparison: start vs enable · Download PNG · Open card page

Server stuck at GRUB?. Wrong kernel or bad config. Pick an older kernel in menu. Rebuild the GRUB config. — Interview card: Server stuck at GRUB? · Download PNG · Open card page

Why initramfs?. Small temporary root FS. Loads drivers to find disk. Rebuild after driver change. — Interview card: Why initramfs? · Download PNG · Open card page

runlevel vs target?. Targets replace runlevels. multi-user = 3, graphical = 5. See the default target. — Interview card: runlevel vs target? · Download PNG · Open card page

Service fails at boot?. Check status first. Read logs for this boot. Fix, then enable it. — Interview card: Service fails at boot? · Download PNG · Open card page

More cards: Ravi Sir's Cards.

Linux Boot Process Step by Step: BIOS/UEFI, GRUB, Kernel, initramfs and systemd

1. An everyday example: opening a shop in the morning

Why do we need this?

What is it?

How to do it (step by step)

2. Stage 1: BIOS or UEFI firmware

Why do we need this?

What is it?

How to do it (step by step)

3. Stage 2: GRUB, the boot loader

Why do we need this?

What is it?

How to do it (step by step)

4. Stage 3: the kernel starts

Why do we need this?

What is it?

How to do it (step by step)

5. Stage 4: initramfs

Why do we need this?

What is it?

How to do it (step by step)

6. Stage 5: systemd and targets

Why do we need this?

What is it?

How to do it (step by step)

7. Reading boot logs and finding slow boots

Why do we need this?

What is it?

How to do it (step by step)

8. Real examples from many industries

Why do we need this?

What is it?

How to do it (step by step)

9. Common errors and how to fix them

Why do we need this?

What is it?

How to do it (step by step)

10. Troubleshooting step by step

Why do we need this?

What is it?

How to do it (step by step)

11. Lab 1: Map the boot of your own server

Why do we need this?

What is it?

How to do it (step by step)

12. Lab 2: Break and fix fstab safely (test VM only)

Why do we need this?

What is it?

How to do it (step by step)

13. Frequently asked questions

14. Interview questions and answers

Beginner level

Mid level level

Senior level

Practice Tasks

Easy Is the target active?

Easy Find slow boot units

Easy Critical chain of a unit

Easy Kernel version string

Easy List failed units

Easy Repair target

Medium Set multi-user target

Medium Errors from this boot

Medium Logs from two boots ago

Medium Rebuild GRUB on Ubuntu

Medium Rebuild initramfs (Ubuntu)

Medium Rebuild initramfs (RHEL)

Medium Test fstab without reboot

Medium Kernel errors only

Medium Keep booting with a missing disk

Hard Rebuild GRUB on RHEL

15. Related chapters

Revision, comparison and interview cards