hesela.devData durability
Data durability is the ability of a storage system to preserve data over a stated period and within a specified failure model.
Preserving data and maintaining access are different goals. An intact copy can outlive a failed access path without being immediately reachable by its host. The tolerated failures and recovery mechanism must be specified. [1]
Object-storage durability targets can be published alongside separate availability and zone-coverage properties. A target is not an observed loss sample, and equal target percentages do not establish equal disaster coverage. [2]
When comparing estimates, record whether the event is first loss or lost volume, and whether the denominator is objects, bytes or systems. The companion analysis separates those questions from access-error arithmetic.
Relations
- related
- Storage availability, Mean time to data loss (MTTDL), Normalized magnitude of data loss (NOMDL), Failure domain
Citations
- SNIA: NVM PM Remote Access for High Availability, May 2019, section 4.1, pp. 13-15 (non-normative whitepaper)
- AWS S3 User Guide: storage-class design targets and zone coverage
Sources & caveats
- Conceptual definition; not a claim of measured reliability. The SNIA source discusses persistent-memory recovery, not a certification for all storage systems.