August 1, 2026
Drive drama: backup or bust?
P[drive failure]: how reliable is your NAS?
One person tried to make home storage last a lifetime, and the comments instantly turned into Backup Court
TLDR: The article is about building a home file setup that can protect irreplaceable memories for decades by keeping multiple copies in different places. Commenters agreed safety matters, but fought over the method: some said the math was overkill, while others repeated the golden rule that storage is not the same thing as backup.
A would-be data immortal went full doomsday planner for family photos, mapping out a home storage setup meant to survive for roughly 50 years. The plan: four hard drives, split into two mirrored pairs, with one copy kept far away at dad’s house and synced automatically. The goal is simple even if the math gets wild: keep precious stuff like photos, messages, backups, and tax records from vanishing into the digital void.
But the real fireworks were in the comments, where the community basically slammed the big red "you are overthinking this" button. The loudest reaction came from the classic internet safety brigade: “A NAS is not a backup device” — a blunt reminder that a fancy box full of drives is still not the same as a true backup plan. Another commenter side-eyed the author’s deep probability calculations, arguing that with only four drives, real life is too chaotic for neat formulas. Translation: one person’s drive lasts 20 years, another’s dies before the warranty vibes even wear off.
Then came the no-nonsense crowd with the digital-prepper mantra: use mirrors for safety, follow the 3-2-1 rule (three copies, two kinds of storage, one kept elsewhere), and stop spending extra money if the files are easy to replace. And in peak comment-section comedy, one person skipped the storage debate entirely just to say they liked the site design. Honestly? In a thread about data death, that may have been the calmest take of all.
Key Points
- •The article proposes a personal NAS design intended to preserve important data over an approximately 50-year period with very low risk of loss.
- •The planned setup uses four 8 TB drives arranged as two separate Btrfs RAID1 mirrors, with one local copy and one offsite copy.
- •The author recommends using Btrfs RAID1 instead of dup because dup can place both copies of data on the same physical device.
- •Btrfs checksumming and self-repair on read are presented as protections against corrupted blocks when a good redundant copy exists.
- •The article distinguishes between full drive failure and sector-level read failures such as latent sector errors or unrecoverable read errors, and describes replacing failed drives with `btrfs replace -r`.