• Sources: primary, discussion
  • Summary: The incident began 2026-09-09 and was marked resolved 2026-10-22 13:34 UTC. Hetzner traced the stalls to low-level commit operations hanging for seconds at a time. It later blamed increased utilisation and I/O pressure, and it moved buckets within HEL1 in the background to fix it.
  • Why it matters: Teams that use Hetzner Object Storage in HEL1 for backups or artifacts should check whether timeouts since September caused failed jobs or partial uploads.