Disaster Recovery

Disaster Recovery

Disaster recovery encompasses the processes, policies, and procedures put in place to restore critical operations for the resumption of business. These critical operations include data access and communications. Disaster recovery spans from simple to complex.

Disaster recovery is defined by two fundamental factors:

"How old is the data I am restoring?" - Recovery Point Objective (RPO) defines the acceptable amount of data loss measured in time. If a disaster occurs and recovery is necessary, what is the acceptable age of the data? An RPO of 24 hours is typical in a nightly-backup scenario. This defines that data, when the restoration process begins, will be no more than 24 hours old.
"How fast can I restore my data?" - Recovery Time Objective (RTO) defines the acceptable amount of time allowed from the identification of a disaster through the disaster recovery process for the full restoration of critical operations. If a disaster occurs and recovery is necessary, what is the acceptable amount of time to restore operation? Defining an RTO of 8 hours would mean once the restoration process begins, critical operations will be online in 8 hours or less.

Defining an RPO and an RTO per critical business operation is the first step in defining a disaster recovery plan and immediately serves to focus the technology requirements to achieve them. For a simple environment an RPO of 24 hours and an RTO of 8 hours can be acheived through a well designed backup to tape strategy. An RPO of 4 hours and an RTO of 1 hour can easily be achieved by leveraging snapshot and replication technologies. An RPO of 0 and an RTO of 0 defines a special case of disaster recovery called business continuance.

Simple (Off-site Tape Rotation)
A simplistic disaster recover plan involves taking the latest tape backup set off-site to a secure location. In order to restore in the event of a disaster, this scenario would require a facility with servers to house the restored data/services, a server running appropriate backup software, and an attached device able to read the off-site media set. Simple disaster recovery plans can fulfill high RPO's of 24 hours and above. The RTO of such a plan is contingent on the speed of the hardware (tape, drive, backup software) and can vary significantly.

Off-site tape rotation often involves contracting third-party companies for off-site tape storage services.

Comprehensive (Off-site Replication)
As the requirements of a disaster recovery plan increase in complexity a more comprehensive solution is needed. With RPO's defined below 24 hours and RTO's defined below 8 hours (as an example) the requirements for the underlying technology increase. Features and functions such as snapshot, replication, and WAN acceleration come into play and are often a requirement in achieving aggresive disaster recovery scenarios. As the requirements increase the complexity and room for human-error increase as well. Product suite solutions providing automation to the disaster recovery infrastructure are key to minimizing human-error.

Off-site replication also provides the option to eliminate third-party contracts for off-site tape storage services.