Many businesses are regulated and are required to provide active
site disaster recovery systems and facilities. For each business application,
there is a defined Recovery Time Objective (RTO) which specifies the maximum
amount of downtime that the application will incur in the event of a site
disaster. In addition to RTO, each application will have a Recovery Point
Objective (RPO), which relates to the freshness of data after a recovery. For an
application to be down for only one second (RTO) and to lose just one second’s
worth of data after a complete site disaster, the most costly and sophisticated
replication, networking and clustering systems must be employed. There are many
IT solutions that are much more cost effective and provide reasonable RTO/RPO
(less than one hour) that would satisfy a great majority of a business’
Disaster recovery service providers can offer a complete set up
of servers, networks and storage, ready to go, to recover from site disaster.
Unfortunately, the cost of these services is out of the reach of most mid-sized
Some companies keep their own hot spare systems (servers,
storage and networks) at a second location for their most business critical
applications. Disk mirroring and replication systems can replicate data from a
primary site, across the metropolitan or wide area network to a second site.
The most commonly deployed site disaster recovery solution
employed involves rolling out cold spare storage subsystems, networks and
servers. Most organizations have days to recover from a site disaster and will
either keep spare systems at a second location or have a service contract with a
provider who will provide spare systems within four hours or less from event
Recovery data for these spare systems typically comes from
backup tapes. Many companies are moving to disk-based backup repositories to
reduce the recovery time after a disaster to minutes instead of hours to days.
Also, a disk-based backup repositories can replace offsite tapes and the
management overhead of offsite tapes.