Recovery Point Objective (RPO) and Recovery Time Objective (RTO) are the two numbers that define a backup and disaster-recovery policy. RPO is the maximum acceptable amount of data loss measured in time — if your nightly backup runs at 02:00 and the disk dies at 14:00, your RPO is 12 hours. RTO is the maximum acceptable downtime between failure and the service being usable again — restoring a 200 GB database from a cold S3 backup might take 2 hours, so your RTO is at least that. Lower numbers cost more.

In a self-hosting context

When you self-host an OSS alternative on a single VPS, your default RPO and RTO are dictated by your backup cron and how fast you can spin up a fresh VPS and restore. Backup tools like restic, Borg, or any S3-compatible target (MinIO, Garage) let you tune RPO down to minutes; pairing a VPS snapshot policy from any provider in the VPS providers compared comparison lets you tune RTO to under 30 minutes. Highly available setups push both numbers toward zero at the cost of running redundant hardware.