• From power-on to userspace
• Firmware: BIOS vs UEFI, Secure Boot
• Bootloaders: GRUB, systemd-boot
• Kernel early boot and initramfs
• systemd architecture and units
• Targets, dependencies, and ordering
• Troubleshooting boot and services
• Exercises

1. Power-on reset → firmware initializes hardware.
2. Firmware locates boot device and loads bootloader.
3. Bootloader selects and loads kernel (+ initramfs) and passes cmdline.
4. Kernel initializes subsystems, mounts rootfs, executes PID 1 (init/systemd).

• UEFI replaces legacy BIOS with a standardized environment and GPT partitioning.
• Secure Boot validates boot chain signatures to prevent unauthorized loaders.
• NVRAM variables store boot entries; ESP (EFI System Partition) holds loaders.

• GRUB: flexible, modules for filesystems, menus, rescue shell.
• systemd-boot: simple UEFI loader reading entry files from ESP.
• Kernel command line controls early boot (e.g., root=, quiet, systemd.unit=, rd.luks.name=).

// Parse kernel cmdline (conceptual)
function parseCmdline(cmd){
  return Object.fromEntries(cmd.split(/\s+/).map(kv=>kv.split('=')));
}

• Initramfs is a cpio archive containing minimal userspace to assemble rootfs (LVM, RAID, LUKS).
• udev populates /dev; early mounts of /proc and /sys; switch_root to final rootfs.
• panic/recovery paths: dracut/emergency shell for diagnostics.

• PID 1 manages processes via declarative unit files: service, socket, mount, automount, timer, target, slice, etc.
• cgroups v2 integration: each unit gets a cgroup subtree for resource control and accounting.
• journald provides structured logs; logind manages sessions; networkd/resolved optional.

# Example: /etc/systemd/system/myapp.service
[Unit]
Description=My App
After=network-online.target
Wants=network-online.target

[Service]
ExecStart=/usr/local/bin/myapp --port=8080
Restart=on-failure
User=myapp
Group=myapp

[Install]
WantedBy=multi-user.target

• Targets group units (e.g., basic.target, multi-user.target, graphical.target).
• Ordering vs requirement edges: After/Before vs Requires/Wants; avoid cycles.
• Activation: socket, path, device, and timer units start services on demand.

# Timer + service pair
# myjob.timer
[Timer]
OnCalendar=*:0/5
Unit=myjob.service

# myjob.service
[Service]
Type=oneshot
ExecStart=/usr/local/bin/myjob

• Use systemd.unit=rescue.target or emergency.target for recovery.
• journalctl -b, -u, and -k for boot and service logs; systemctl list-dependencies for graphs.
• Kernel params: systemd.log_level=debug, systemd.mask=UNIT to isolate issues.

1. Create a systemd service + timer that runs a maintenance job and logs output to journal.
2. Build a minimal initramfs (dracut or mkinitcpio) and boot a VM from it; add an emergency shell.
3. Switch default target and trace boot ordering with systemd-analyze blame/critical-chain.