<< back Page 3 of 3

NoSQL, NewSQL, and Emerging Blended Enterprise Data Environments

When it comes to high-level analysis, traditional relational databases are well-suited for the tasks, and are designed for this very purpose. Extract, transform, and load (ETL) processes were designed to support RDBMS environments, which enable deep digging into datasets. Even Hadoop data can be loaded into relational data warehouses. However, NoSQL may be a preferable approach for real-time analytics.

BI and Analytics Tools Compatibility

Most enterprises are now laden with a variety of business intelligence and analytics tools and platforms. However, many BI tools, which were introduced in the RDBMS world, are still not compatible or ready to support NoSQL environments. The ability to compete on analytics requires the capability to quickly tap into data sources or data streams, and deliver insights to decision makers or applications as they are needed. NoSQL data environments—with technology that is faster, more responsive, and more lightweight than traditional RDBMSs—offer a compelling platform for today’s burgeoning analytics needs.

NoSQL and NewSQL environments enable the capture and analysis of unstructured data types in a relatively quick and inexpensive manner. For example, sensor- or machine-generated data has actually been present in many enterprises for quite some time—but to attempt to bring this type of data into relational database or data warehouse environments was prohibitively expensive.

Emerging Requirements in Security, Skills, and Self-Service

There’s no question that today’s corporate data stores are constantly under attack, or are vulnerable to potential breaches. This cuts right to the viability of businesses—sizable data breaches have major repercussions in terms of liability, regulatory sanctions, and brand image. RDBMSs are considered highly secure, while NoSQL security is seen as relatively immature—though this market is evolving quickly.

In addition, while relational database management skill sets have been around for quite some time, skill sets for NoSQL/NewSQL still have to reach critical mass. These emerging data environments call for investment in new skill sets beyond the relational model, and data managers and professionals need to play a consultative role in understanding what types of technologies and platforms should be advanced, and what parts of enterprise systems need to be modernized, replaced, or virtualized.

At the same time, business users accustomed to the rapid delivery of both search and analytics results from sites such as Google or Bing are increasingly frustrated with the slow pace of enterprise analytics systems. This gets especially frustrating if it involves sending requests for reports to IT and waiting for eventual delivery. However, self-service analytics—powered by NoSQL—opens up the data kingdom, enabling rapid and frequent queries as they are needed, enabling users across organizations to build their own reports, dashboards, scorecards, analysis, ad hoc queries, and planning models—working with corporate data that is rapidly and readily obtainable and downloadable from back-end or mission-critical systems.

The New Blended Database Environments

Achieving a productive and forward-looking co-existence between various database environments requires collaboration and vision. With the continuing rise of unstructured data as important information sources, enterprises are reevaluating the way they are architecting, storing, and analyzing their data. This information ultimately has to be put to productive use, and NoSQL and NewSQL databases provide new ways of meeting the big data surge, virtually ensuring that every imaginable data type—no matter how much there is—can be examined and analyzed by enterprises.

Blended database environments—incorporating the best qualities and strengths of both relational and NoSQL/NewSQL technologies—are core to next-generation data center architecture. The key is to foster an environment of coexistence between relational database systems and unstructured data environments.

<< back Page 3 of 3


Subscribe to Big Data Quarterly E-Edition