Backup Retention Policy

How a Retention Policy Sets Your Restore Window

A backup retention policy is the rule that sets how long each backup is kept before it is deleted or replaced. It determines how far back you can restore a site, which matters as much as how often copies are made.

A retention window is measured in restore points, the individual copies you can roll back to. If you keep seven daily copies, the oldest restore point is about a week old. Anything that happened before that point cannot be recovered from backups, even if the problem was never noticed.

A common structured way to set these windows is covered in grandfather-father-son rotation.

Why Short Retention Fails

Short retention works well for sudden, obvious failures, such as a server that stops responding. It fails for slow or hidden problems.

• Hacks with delayed discovery. Malicious code may sit unnoticed for weeks. If every kept copy was made after the infection, none of them is clean.
• Gradual corruption. A database fault or a faulty plugin may quietly damage records over time. The damage may only become visible after the oldest good copy is gone.
• Deletion discovered late. Content removed months ago may be needed later for a legal or customer request, and a short window cannot bring it back.

Longer retention costs storage, so the goal is a balance. Some businesses also have legal or accounting obligations to keep certain records for set periods. Keeping data for years is a different job, explained in backup vs archive.