In a storage cluster at hosting provider Hetzner’s Nuremberg data center, about 1,500 snapshots were lost on Tuesday that could not be restored despite redundancy. Many customers of the Hetzner cloud service were affected. A curious failure of several hard drives is to blame for this.

Hetzner confirmed the process is going online. According to this, the data is actually protected from errors by error coding and is distributed over multiple hard drives via Ceph. This scenario kept the saved snapshots intact and available even if two disks failed.

combination of multiple failures

A company spokesperson also pointed out that the extremely rare case that two hard drives failed at the same time was not attributable to the data failure alone. However, during the rebuild of everything, another hard disk failed, causing the loss of data from 1500 snapshots.

The Hetzner Company immediately notified affected customers by email of the loss of data and apologizes for the inconvenience. And you take this incident as an opportunity to check the data backup process and rectify it if necessary.

The fact that the incident was not adequately communicated externally by the company was met with outrage online. not that either Twitter accountNo status page The provider informed about the data loss.

(mavi)