Having backups isn’t the same as being able to recover. A disaster recovery plan defines how much time and how much data your business can afford to lose, and designs the infrastructure around that.
Two numbers that define everything
- RPO (Recovery Point Objective): how much data you can lose, measured in time. A one-hour RPO means losing up to the last hour of information is acceptable.
- RTO (Recovery Time Objective): how long the service can be down before it’s back up.
The lower these values, the higher the cost. That’s why it pays to define them together with business teams.
The four most common strategies
- Backup and restore: the most affordable option. Data is copied regularly and the infrastructure is recreated when needed. RTO measured in hours.
- Pilot light: data is replicated and critical components are ready but switched off. RTO in the tens of minutes.
- Warm standby: a scaled-down version of the environment always runs in another region and scales up when something fails.
- Multi-region active-active: the service runs in several regions at once. Near-zero RTO, at the highest cost.
Tools that help
AWS Backup centralizes and automates backups of databases, volumes and file systems, with retention policies and cross-region copies. AWS Elastic Disaster Recovery continuously replicates servers so you can recover them in minutes.
The most forgotten part: testing
A plan that has never been tested is a hypothesis. Schedule regular restore drills and document the results. At Nubilink we design and test recovery strategies as part of our managed services.