RAID Is Not a Backup: The Complete Guide to Data Survival

“RAID is not a backup” is a mantra in the homelab community, and our episodes explored exactly why — along with what a real backup strategy actually looks like.

RAID: The Foundation

• RAID Demystified started with the basics. RAID levels, parity calculations, rebuild times, and the critical distinction between redundancy (protection against drive failure) and backup (protection against data loss). The hosts made the key point: RAID protects uptime, not data. A ransomware attack or accidental deletion wipes a RAID array just as effectively as a single drive.

When Things Go Wrong

• Server Resurrection was a war story — a dead motherboard, degraded arrays, and the recovery process. The lessons were practical: label your drives, document your array configuration, and keep a cold spare controller. ZFS Decoded went deeper into ZFS-specific recovery, covering scrubs, resilver operations, and the surprisingly non-trivial process of importing a ZFS pool on new hardware.

Building Resilience

• RAID is Not a Backup (the episode that earned its title) laid out the 3-2-1 strategy: three copies, two different media types, one offsite. The hosts discussed snapshot-based backup with tools like Sanoid/Syncoid, versioned offsite sync, and the role of object storage services as a cheap offsite tier.

The Frontier

• IPFS vs. The Cloud explored decentralized storage as a backup strategy — content-addressed, distributed, and theoretically permanent. The reality is more nuanced: IPFS is excellent for data you want to be publicly available and resilient, but it’s not a drop-in replacement for traditional backup.

• The Billion-Year Backup went to the extreme end: DNA storage, quartz glass, lunar archives, and the question of whether any storage medium can truly last forever. The answer is humbling — our best digital archives are fragile compared to a clay tablet.

• Beyond Backups zoomed out to infrastructure-level redundancy: what happens when the backup site also fails? Geographic distribution, failure domain analysis, and the economics of diminishing returns in resilience planning.

---

The message across all seven episodes: data you haven’t backed up is data you’ve already lost — you just don’t know it yet. Build your strategy around the assumption that every single component will eventually fail.