Claude spotted failing HDD from SMART data, insisted on backup — drive died right after
Disastrous_Friend1 · reddit · 2026-08-22
A Reddit user working with Claude (Opus 5-extra) on a project noticed the model unpromptedly flagged unusually slow disk writes, dug into SMART data and journals, and found bad sectors climbing from 16 to 216 — a strong signal of imminent drive failure. The user brushed it off, but Claude pushed back and insisted on an immediate backup. During copying, 20% of the previous week's files turned out corrupted and the prior day's writes were 100% corrupted; right after the backup finished, the drive stopped responding entirely. Claude then used git history and session history to repair the corrupted project files.
More from coding & agent
- Engineer: my job went from silently thinking at a computer to constantly talking to it — KevinNaughtonJr · 2026-10-16
- Vercel Goes Full Backend: Python GA, Containers, and Daemons — cramforce · 2026-10-16
- WorldMemArena Accepted at NeurIPS: Memory Storage ≠ Memory Use — xwang_lk · 2026-10-16
- Vercel Ship Demos Container Deploys for Go, Ruby, Java and More — edwin · 2026-10-16
- TRL adds day-0 support for fine-tuning Muse Spark Open's 816B MoE with LoRA on 64 H100s — QGallouedec · 2026-10-16
- Databricks engineers land impactful PRs on day 2 with AI dev tools, vs week 3 before — Yuchenj_UW · 2026-10-16