0 / 15 lessons — 0%
Lesson 14 / 15
The boot process & recovery
Knowing what's supposed to happen between power-on and a login prompt is what makes it possible to tell where a broken boot actually went wrong.
Each stage hands off to the next. Knowing which one failed tells you which tool to reach for.
| Stage | Job | Symptom if it fails here |
|---|---|---|
| BIOS/UEFI | Initializes hardware, finds a boot device | No display output at all, no beep |
| Bootloader (GRUB) | Presents kernel choices, loads the selected one | Stuck at a GRUB menu, or "no bootable device" |
| Kernel | Initializes hardware drivers, mounts the root filesystem | Kernel panic messages, boot freezes early with driver errors |
| systemd (PID 1) | Starts every configured service, in dependency order | Boots but hangs, or drops to an emergency shell |
# after a failed boot, see exactly why systemd got stuck systemctl --failed # which units failed to start journalctl -b -p err # errors from the most recent boot attempt systemctl list-dependencies # what's supposed to start, and in what order
Dropped into an emergency shell or rescue mode (often from a bad fstab entry, a corrupted filesystem, or a failed critical service)? That's a minimal environment for fixing exactly the problem that's blocking a normal boot — usually editing a config file or running a filesystem check, then rebooting.
Try it yourselfRun
systemctl --failed on any real machine. An empty list is genuinely reassuring; anything listed there is worth a look even if the machine otherwise "seems fine."