Glossary
Secondary storage
Secondary storage holds data that is not serving live production: backups, snapshots and replicas kept for recovery, archives kept for compliance, and files that are rarely opened. It sits behind primary storage, and although it gets less attention, it holds the copy an organization restores from after a failure or attack.
What lives on the secondary tier
| Data | Why it is kept | How often it is read |
|---|---|---|
| Backups | To restore systems and files after loss or damage | Rarely, then urgently and in bulk |
| Snapshots and replicas | Fast recovery of recent versions or whole systems | During incidents and tests |
| Archives | Legal, regulatory or research retention | Occasionally, often years later |
| Inactive files | Still needed but seldom opened | Now and then, by individual users |
Primary storage serves applications in real time and is built for low latency, usually on flash. Data reaches the secondary tier through backup jobs, replication and storage tiering policies, and flows back only when something has to be restored or retrieved. The coldest part of it, read perhaps once in years, is often kept on cold storage.
Attention and importance run in opposite directions
Production systems get the fast media, the monitoring and the budget. The copy that recovery depends on often gets whatever was cheapest per terabyte. If primary storage is lost, though, the secondary copy is the business. It is also a first target for ransomware operators, who often delete or encrypt backups before they encrypt production, because a working backup removes their leverage.
What matters more than cost per terabyte?
- Restore speed. How fast data comes back when a whole environment is recovered at once, covered under restore throughput. A tier that is cheap but slow to read can turn a recovery of hours into one of days.
- Immutability. Whether copies survive someone holding a stolen administrator login. Retention locks enforced by the storage itself, as in an immutable backup, answer this; settings held only in backup software do not.
- Growth. Retention periods and data growth mean the tier expands every year, so how it scales matters more than its starting size.
- Cost. Still real, and weighed after the other three. Many teams keep a fast copy for quick restores and a lower-cost one for long retention, with object storage increasingly used for both.
Frequently asked questions
Is secondary storage the same as backup storage?
Backup is the largest part of it, but the tier also holds replicas, archives and inactive files. Secondary storage is defined by role, not by the software that writes to it.
Does secondary storage have to be slow?
No. Slow media is a cost choice. Recovery targets can require a secondary tier that reads at high throughput, which makes restore speed a design requirement.
What is tertiary storage?
Usually offline media such as tape that has to be mounted before it can be read. Some definitions fold it into secondary storage, others treat it as a separate, slower layer.