Strategies vary by Recovery Time Objective (RTO) and cost, from slowest to fastest recovery:
- Backup and Restore
- Most cost-effective, simplest to implement.
- Data is backed up (e.g., Amazon S3, Amazon S3 Glacier).
- Resources are provisioned and restored after a disaster.
- Highest RTO and RPO.
- Pilot Light
- Minimal critical infrastructure always running (e.g., core databases, essential services).
- Additional components launched on demand during recovery.
- Lower RTO than backup & restore, but still requires scaling up.
- Warm Standby
- Fully functional but scaled-down copy of the environment running at all times.
- Can be scaled quickly to production size during a disaster.
- Balances cost and recovery speed.
- Hot Site / Multi-Site Active-Active
- Full production-scale environment running in another AWS Region or site.
- Traffic can be shifted instantly in case of failure.
- Lowest RTO and RPO, but highest operational cost.