Glossary

Disaster recovery plan

A disaster recovery plan is the written procedure an organization follows to restore IT systems when its main environment cannot be used. It names who can declare a disaster, the recovery targets for each system, which copies to restore from, the order of restoration, where recovery runs and how operations return.

What the plan records

  • Scope. The systems and sites covered, and those explicitly left out.
  • Authority. Who can declare a disaster and approve a failover, agreed before an incident.
  • Targets. Recovery time and recovery point for each system or tier, taken from the business impact analysis.
  • Sources. Which copy each system is restored from, and how it is known to be clean.
  • Order. The restoration sequence, by dependency.
  • Recovery site. Where recovery runs, whether a hot, warm or cold site, with access and capacity details.
  • Contacts and contracts. People, suppliers and support agreements, with numbers that work outside the office.
  • Return. How operations move back, covered under failback.

Restore order follows dependencies

Systems cannot all start at once. A plan that lists them alphabetically or by owner produces a recovery that stalls on a dependency nobody started. The usual sequence is:

  1. Network and connectivity at the recovery site.
  2. Identity, directory and DNS.
  3. Storage and databases.
  4. Applications, in the order their dependencies allow.
  5. User access and external connections.

Once the sequence is stable, orchestration can run it as a single workflow.

Three gaps that make plans fail

The first is a plan, password vault or key manager that lives on the systems being recovered, so it disappears with them. The second is naming a replica as the source when a clean copy is needed: a replica holds whatever production held, including encrypted files, so the plan has to point to a copy that predates the damage, such as an immutable backup. The third is an application added after the plan was written, which is the one left behind. NIST's SP 800-34 treats plan maintenance as part of the plan for this reason.

Frequently asked questions

What is the difference between a disaster recovery plan and a business continuity plan?

A disaster recovery plan restores IT systems and data. A business continuity plan keeps the organization running, covering people, facilities, suppliers and manual workarounds, and the disaster recovery plan sits inside it.

How often is a disaster recovery plan updated?

Common practice is to review it after every significant change to systems or staff and to exercise it at least once a year. A plan that is never exercised is usually out of date before anyone notices.