Two years ago, my colleague Bruce Momjian wrote that relational databases would adapt to emerging database demands and incorporate the capabilities that emerging standalone NoSQL solutions provided. He has been proven correct with the expanded capabilities in Postgres alone with recent releases adding greater scaling, non-durable tables and JSON/JSONB capabilities. Today, you don't need a standalone document database solution if you have Postgres. And recent research shows the new technology comes just in time, as many database professionals admit they struggle with managing standalone NoSQL solutions in general. A commissioned study conducted by Forrester Consulting on behalf of EnterpriseDB, published in January 2015, presents a case study for the evolution of relational database management systems. The study, Relational Databases are Evolving to Support New Data Capabilities, found that the majority—78%—of database decisions makers wanted one solution that could handle relational and NoSQL data types. The study finds that relational databases are evolving to address the needs of end users seeking to link unstructured and structured data types and that decision makers should look to invest in these solutions. EDB’s Postgres Plus Advanced Server, for example, addresses these needs with such capabilities as support for unstructured data types, non-durable tables, tools for large-scale data loads, and integration technologies that connect standalone NoSQL solutions with Postgres. Consider some of the findings from the study:
- Developers’ NoSQL Decisions Carry a Cost. More than half—52%—of respondents said they were unable to prevent developers from deploying new applications on separate NoSQL databases, while 42% admitted they struggle to manage the NoSQL databases deployed on their infrastructures. In addition, 30% reported that data stored in NoSQL solutions were creating data siloes in their businesses.
- Decision-makers want to link unstructured and structured data. More than one-third of respondents—36%—said they would like to link their unstructured data with their structured data all or most of the time. Almost two-thirds, 60%, said they sometimes found a need. Just 4% said they never have the desire.
- Traditional data still holds greatest strategic value. Despite the growing buzz around social, mobile and other new data types, the most important for business strategy continue to be those with longstanding value. Of the respondents, 85% said planning, budgeting and forecasting data was most important to overall business strategy while 72% said transactional data from business apps. This data is typically made more valuable by applying relational qualities, supporting the continued use of a relational database as the corporate standard.
That said, database decision makers recognize new data types play an important role as well, supporting the need for database management systems that can support unstructured data types. Of the respondents, 58% said unstructured internal data was important, 52% said scientific data, 36% said video, imagery and audio and 33% said consumer mobile and device data.
- Decision-makers want one database to support all data types. Close to half—42%—of respondents want to integrate NoSQL databases with relational and just over one third—36%—want to store both structured/unstructured data in their standard database. While end users have not yet determined how they want to achieve it, the goal for this 78% majority is clearly to find a way to support both kinds of data with one database.
EnterpriseDB has been championing the notion that NoSQL should reference “Not Only SQL,” and that organizations need a database management system that supports structured, semi-structured and unstructured data. Based on industry research and what database decisions makers have to say, we’re not alone. Review the Forrester report here or contact us to learn more about how Postgres Plus Advanced Server can provide NoSQL capabilities for your enterprise.
Marc Linster is Senior Vice President, Products and Services, at EnterpriseDB.