Simple definition
A full backup covers all data selected for a protection point. It can restore that scope without replaying an earlier incremental sequence. This independence concerns the included data; it does not prove that every dependency needed by an application has also been protected.
Technical definition
An active full can read source data again, while supported products may build a synthetic full from existing backup files. Synthetic processing can reduce source reads but still places work on the repository. Compression and deduplication affect physical consumption without by themselves changing the logical protection scope.
What is it used for?
Establish a chain baseline, simplify some restore operations and reduce reliance on older recovery points. Choose frequency according to backup duration, retention, acceptable data loss and the time needed to recover a usable service.
Practical example
A team keeps a weekly full and daily incrementals. To recover Wednesday’s state, it checks the base and the additional files required by the product’s backup method. It then validates the application, accounts, certificates and dependencies rather than merely opening a restored file.
Common issues
- An exceeded backup window or undersized repository.
- Stored data without the required application-consistent state.
- Missing encryption keys, catalog information or recovery dependencies.
Operational checks
Check what the job actually includes, the type of full created and its processing messages. Distinguish bytes read, logical protected size and physical repository use before treating a size difference as a fault.
- Confirm the selected data and applications.
- Review consistency, retention and repository capacity.
- Verify access to keys and recovery information.
- Perform an isolated restore and measure time until the service is usable.
What to distinguish
| Element | Role |
|---|---|
| Active full | Reads source data according to product behavior. |
| Synthetic full | Builds a full point from repository data. |
| Incremental | Protects changes within the backup chain. |
| Recovery test | Measures whether a usable service can be recovered. |
Limits and precautions
A successful job status alone does not prove service recoverability. Maintain a protected copy independent of the production failure domain. Before deleting an older full, validate dependent recovery points and the product’s retention rules; deleting a chain base can invalidate later points.
Official documentation
Related terms
Key takeaway: A full backup protects the entire selected scope and provides a recovery point or a base for a backup chain.