Traditionally, relational database have been
the number one choice for storing structured data in enterprise business
applications. Relational data stores have been widely adopted and are often
thought of as the only alternative for data storage accessible by multiple
clients. There have been other approaches over the years, such as Object
and
XML databases, but these technologies never came close to the market share
held
by RDBMS. Instead, these types of innovations simply became absorbed by
newer
generations of relational database management systems.
With the rise of cloud computing, a "one-size fits all" mentality has
emerged concerning data store
architectures, leading to a new type of data store commonly labeled with
the term of "NoSQL" (or "Not only SQL").
NoSQL data stores can be categorised as tabular/columnar, document, graph
and key/value store databases, each optimised
to handle certain kinds of data processing requirements.
The main driver for the creation of NoSQL data stores has been the
popularity of
"Web-scale" data at large, global Web sites and services, such as Amazon,
Google, Yahoo!, and Facebook. Recently, predictive analytics,
voice-of-the-customer, fraud, and other BigData use cases have surfaced to
further accelerate the demand for NoSQL.
But are NoSQL databases only useful in BigData scenarios? Or should they
be positioned as an alternative to relational
databases for persistence in an N-Tier architecture?
This session presents the most popular NoSQL data storage engines,
discuses the factors to consider with the
potential tradeoffs of imposed by NoSQL, and demonstrates how the concept
of data virtualization can help create an abstraction layer that hides the
complexities of the underlying data sources and by that provides a unified
view of enterprise data which can also be used directly for providing Data
Services
in a Service-Oriented Architecture.