Home  ›  Glossary  ›  Forward Incremental Backup

What is forward incremental backup?

Forward incremental backup is a backup method that starts with a full backup and adds an incremental after each job, each holding the changes since the previous one. Restore points move forward in time along the chain, and the newest point is restored from the full plus every incremental taken since it.

How is a forward chain built?

A full backup starts the chain. Each later job writes a new file holding only what changed since the job before, and existing files are never modified. A new full, read from production or built as a synthetic full, starts the next chain. A weekly full with daily incrementals is a common pattern and gives chains of seven restore points.

Why do chains stay longer than the retention setting?

An incremental is useless without the full and every incremental before it, so a chain can only be deleted as a whole once its newest restore point has expired. With 14 days of retention and a full every Sunday, a chain stays until its Saturday incremental passes 14 days old. At that peak, storage holds 20 days of restore points rather than 14, and capacity planning has to allow for the difference.

What does a restore depend on?

Restoring the newest point means reading the full and applying each incremental in order. A missing or damaged file anywhere in the chain breaks every restore point after it, which is the main exposure of the method. Periodic fulls cap that exposure at the length of one chain: the shorter the chain, the fewer files any restore relies on.

Why does it suit immutable storage?

With periodic fulls, every file is written once and never altered. That matches storage that locks data against change, such as S3 Object Lock, with one condition: the lock on the full has to last as long as the newest incremental that depends on it. The incremental forever variant, which merges old changes into the full, alters the base and needs the backup software to handle the merge without touching locked data.

Frequently asked questions

How does forward incremental differ from reverse incremental?

Forward keeps the oldest restore point as the full and adds changes after it. Reverse keeps the newest restore point as the full and stores earlier states behind it.

Is an active full or a synthetic full better for starting a new chain?

An active full reads production again and starts from clean source data. A synthetic full avoids that read but carries forward any damage already in the old chain.

What sets the length of a forward chain?

The interval between fulls. Longer chains save capacity and production reads; shorter chains reduce how many files each restore depends on.