Google Drive misplaces months' worth of customer data

Original Article ↗ Hacker News Discussion ↗

Reports that Google Drive has lost months of user files are fueling broader worries about relying on big cloud providers for long‑term data storage. Commenters debate what ownership and rights really mean under Google’s terms of service, how much legal recourse users have—especially on free tiers—and whether data loss stems from backend failures or sync client bugs. Many conclude that critical files should never exist only in one cloud account, recommending independent backup strategies (rclone, NAS, Synology, CubeBackup, multi‑provider redundancy) and treating “the cloud” as one component in a wider backup plan rather than a single source of truth.

Scope of the incident & reliability concerns

  • Commenters report missing Drive files, reverted folder structures, and reappearing deleted files; some also report Gmail messages and Google Photos items vanishing over wide date ranges.
  • Several say they’d long suspected silent data loss or random content deletion in Google Docs/Drive and were previously dismissed by support.
  • Google’s official status dashboard shows no incident; many see first‑party status pages as untrustworthy marketing rather than diagnostics.

Data ownership, rights, and liability

  • Extended debate over “not your drive, not your data”:
    • Some argue legal ownership is meaningless if a provider can delete data with no effective remedy, especially on free tiers with liability capped at $0.
    • Others counter that you retain IP rights, can in theory sue for damages, and that liability caps don’t erase all tort claims (though enforcement is costly).
  • Analogies to landlords, burglars, and cloud contracts illustrate that practical recourse, not abstract ownership, is what matters.

Backup expectations and practices

  • Strong consensus: never trust a single cloud provider; if you have only one copy in the cloud, it’s not a backup.
  • Many describe 3‑2‑1–style setups:
    • rclone to mirror Drive/OneDrive to Dropbox, S3, Backblaze B2, Glacier, or NAS.
    • Google Takeout exports on a schedule, sometimes scripted and streamed via encryption tools to object storage.
    • Dedicated tools (CubeBackup, Synology ActiveBackup, Duplicati, Restic) backing up Workspace accounts or personal Google data.
  • Some stress regularly testing restores and using checksummed filesystems (btrfs/ZFS) to detect bit‑rot and corruption.

Alternatives and self‑hosting

  • Suggestions range from Dropbox, Mega, Backblaze B2, and AWS S3 to self‑hosted Nextcloud/Synology/Immich/Piwigo plus offsite disks.
  • Several emphasize “uncorrelated redundancy”: multiple providers + local copies, not betting everything on one cloud.

Speculation on root cause

  • Guesses include:
    • A stale replica promoted to primary, causing a 6‑month rollback and complex merge conflicts.
    • Bugs or flaws in desktop sync clients overwriting locals with old server state.
    • Backend storage glitches returning bad pointers or misfired “old file deletion” processes.
  • No clear technical root cause is established in the thread.

Product & UX frustrations

  • Many criticize Google Drive’s file browser and search; others say building Drive at scale is genuinely hard but paid tiers still imply a duty of care.
  • Comparisons to OneDrive/SharePoint/Teams and Workspace vs personal Google accounts highlight broader dissatisfaction with big‑vendor collaboration stacks.