Module 4 · Storage, Backups & Virtualization

Storage is the most under-taught topic in all of infrastructure — right up until the day data is lost, when it becomes the only topic. This module fixes that. You’ll learn how disks and filesystems really work, build the backup discipline that separates professionals from hobbyists (including the graded exercise everyone should do at least once: delete something important and restore it), and then meet virtualization — the abstraction that everything in modern infrastructure sits on top of.

By the end, your Module 2 server won’t be a single fragile machine anymore. It’ll have redundant storage, tested backups, and a hypervisor running your services as VMs you can snapshot, break, and roll back without fear.

What you need

• Your Module 2 server (the micro PC), now with a network address from Module 3.
• At least one extra disk — a second SSD, a spare 2.5“ HDD, or even a USB drive — for the storage and RAID labs. Two extra disks is better (for a mirror). Used drives are fine; you’re learning, not storing anything precious yet.
• An external USB drive to act as a backup target (the “offsite-ish” copy in 3-2-1).
• See the hardware guide Tier 2 for cheap options.

This module writes to disks — and mistakes erase data

Partitioning and filesystem commands can destroy data instantly and permanently, like dd in Lesson 2.1. Do the destructive labs on spare disks with nothing you care about, and confirm every device name before you act. That caution is itself part of the lesson: respect for storage operations is what this module is trying to instill.

The lessons

Lesson	Topic	Time
4.1 · Disks & Filesystems	Block devices, partitions, ext4, LVM, SMART	4–6 hrs
4.2 · Redundancy Is Not Backup	RAID concepts, ZFS, the crucial distinction	4–5 hrs
4.3 · Backups: The 3-2-1 Discipline	restic/borg, 3-2-1, and testing restores	4–6 hrs
4.4 · Virtualization	Hypervisors, Proxmox, VM lifecycle, VMs vs containers	5–7 hrs
Labs	The five graded exercises	8–12 hrs

Total: roughly 30–40 hours, or 3–4 weeks part-time.

The mantra of this module

Two sentences to carry for the rest of your career:

RAID keeps you running through a disk failure. Backups save you from deletion, ransomware, and yourself. They are different problems, and you need both.

Half of this module exists to make sure you never confuse the two — a confusion that has cost real companies everything.

Checkpoint

• I can partition, format, mount, and resize storage without a tutorial open
• I can explain why RAID is not backup in one sentence to a non-engineer
• My homelab has automated, encrypted, tested backups
• I have performed a timed restore and have the runbook to prove it
• My physical server runs a hypervisor; my services now live in VMs/containers
• I snapshot before risky changes, reflexively

Deliverable

A verified restore demo: a blog post with your backup architecture (diagrammed in Mermaid), your restore runbook, and the receipts — timestamps and terminal output of a real restore drill. Interviewers notice this one, because almost nobody does it. Full spec in Lab 3.

Resources

• Proxmox VE documentation
• restic docs — the backup tool you’ll actually stick with
• Arch Wiki: LVM and ext4 — concept references
• Jim Salter’s ZFS articles (Ars Technica) — the best plain-language ZFS explanations anywhere