Administrator Guide

Synchronous replication use cases
26 Dell EMC SC Series: Synchronous Replication and Live Volume | CML1064
Database replicas distributed in a mixed topology
4.5 Disaster recovery
With data footprints growing exponentially, backup and maintenance windows shrinking along with the cost of
storage, and the impact of downtime gnawing on the conscience of businesses, migrating to online storage-
based data protection strategies is trending for a variety of organizations. Legacy processes which dumped
data to tape were once cost effective and acceptable, but the convergence of key decision-making factors has
prompted a shift from yesterday’s nearline storage to the more affordable and efficient online storage of
today. Data replication within or between sites is the ubiquitous backbone for much more scalable data
protection strategies. With replication in place as the data mover, an assortment of vendor and platform
provided tools and methods can be coupled to replication to form a manual or electronically documented and
reliable recovery process. The SC Series support for multiple replication topologies really comes into play in
the disaster recovery conversation because it adds a lot of flexibility for customers wanting to provide data
protection for multiple or distributed site architectures. Before getting into platform-specific examples, two
fundamental disaster recovery metrics need to be understood as they will be referenced throughout business
continuation planning discussions.
Recovery point objective (RPO): This is the acceptable amount of data loss measured by time or the
previous point in time at which data is recovered from. An RPO is negotiated with business units and
predefined in a disaster recovery plan. In terms of replication, the keys to achieving an RPO target are
choosing the appropriate replication type, making sure replication is current (as opposed to out of date or
behind), and knowing the tools and processes required to recovery from a given restore point.
Recovery time objective (RTO): This is the elapsed recovery time allowed to bring up a working production
environment. Just like RPO, RTO is negotiated with business units and predefined in a disaster recovery plan
and may also be included in a service level agreement (SLA). The keys to achieving targeted RTO may vary
from data center to data center but they all revolve around process efficiency and automation tools wherever
possible. Replication is a quintessential contributor to meeting RTO, especially at large scale.
By leveraging replication, aggressive RPOs and RTOs can be targeted. Data footprint and rate of change
growth may be continuous, but feasible RPO and RTO goals do not linearly diminish as long as the replication