Full, differential, and incremental backups

WHAT THIS TESTS Whether you understand that backup choice is a tradeoff between backup time and storage on one side and restore time and reliability on the other.

A GOOD ANSWER COVERS A full backup copies the entire dataset. It is the simplest to restore, a single operation, but consumes the most space and time to create. A differential backup captures everything that has changed since the last full backup; each successive differential grows larger as more changes accumulate, but a restore needs only the full plus the single most recent differential. An incremental backup captures only what changed since the previous backup of any kind, full or incremental. Incrementals are the smallest and fastest to create, but restoring requires the last full plus every incremental taken since, applied in order, so a single missing or corrupt incremental breaks the chain.

COMMON WRONG ANSWERS Reversing the definitions of differential and incremental, or claiming incrementals are always best while ignoring that their restore chain is longer and more fragile.

LIKELY FOLLOW-UPS How this maps to recovery time objective and recovery point objective; a common rotation like weekly full plus daily incrementals; why you periodically test restores.

ONE CONCRETE EXAMPLE A team runs a full backup every Sunday. With incrementals each weeknight, Thursday's restore needs Sunday's full plus Monday, Tuesday, Wednesday, Thursday backups. With differentials instead, Thursday's restore needs only Sunday's full plus Thursday's differential, simpler to recover but each differential is larger than an incremental. They pick incrementals when nightly backup windows are tight and accept the longer restore.