Understanding Disaster Recovery (容灾) and Backup: Differences, Classifications, and Levels

Disaster recovery (容灾) refers to establishing two or more geographically separated, functionally identical IT systems that monitor each other's health and can switch functions when one site fails due to incidents such as fire or earthquake, ensuring continuous operation.

Disaster‑recovery technology is a component of high‑availability solutions, focusing on mitigating external environmental impacts and providing node‑level system restoration.

There are two main types of disaster‑recovery systems: data‑level (real‑time replication of critical application data to a remote site) and application‑level (a complete backup application system at a remote site that can take over business operations during a disaster).

While both disaster recovery and backup aim to protect data, disaster recovery ensures business continuity by keeping services running, whereas backup primarily safeguards data against loss; disaster recovery is typically an online, real‑time process, while backup is offline and may involve time‑lagged copies.

The article lists six fundamental differences between disaster recovery and backup: (1) disaster recovery sites are physically distant to survive major natural disasters, while backups often reside in the same data center; (2) disaster recovery protects both data and business continuity, backup protects only data; (3) disaster recovery maintains data integrity, backup restores only to a previous point; (4) disaster recovery is an online process, backup is offline; (5) disaster‑recovery data is continuously synchronized, backup data may be stale; (6) disaster‑recovery switchover occurs within seconds to minutes, backup restoration can take hours.

Disaster‑recovery solutions are further classified into three levels:

Data‑level : basic remote data backup; low cost, simple implementation, but application downtime during a disaster.

Application‑level : builds a full replica of the production application at the remote site, providing complete, reliable, and secure services to maintain business continuity; higher cost and complexity.

Business‑level : full‑scale disaster‑recovery covering all IT and infrastructure components, ensuring end‑to‑end business continuity; highest cost and implementation difficulty.

Backup levels are described from Level 0 (no off‑site backup) to Level 3 (active hot‑site with real‑time mirroring). Level 0 offers no disaster‑recovery capability. Level 1 uses local tape backup stored off‑site. Level 2 employs a hot backup site that receives synchronized or asynchronous data copies but does not run applications. Level 3 involves two fully operational data centers that replicate each other; variants include key‑data mirroring only or full zero‑data‑loss mirroring, requiring complex management software and dedicated hardware.

Choosing an appropriate disaster‑recovery and backup strategy requires considering factors such as data volume, distance and bandwidth between primary and backup sites, required recovery time objectives, management overhead, and budget.