Secrets to 50 years of relational database success: innovation, evolution

But keeping pace with the ever-changing nature of data, applications, workloads, and deployment models is no simple task.

John Foley | February 19, 2024


The relational database has reached a golden milestone—it’s been 50 years since IBM launched a project to develop the first prototype relational database based on a seminal paper by E. F. Codd. That was followed by Oracle’s introduction of the industry’s first commercial relational database management system (DBMS), Oracle Version 2, in 1979. These were historic steps on the path to modern data management.

A lot has changed since those early days of enterprise computing. Back then, system memory and disk storage were measured in kilobytes and megabytes, respectively. Today, some large organizations have data estates that comprise exabytes of data—that’s trillions of megabytes.

However, one thing that hasn’t changed is the viability of the 50-year-old relational database model as the preferred way of organizing, administering, analyzing, distributing, and sharing critical business data. In fact, even as other database technologies, including object-oriented, graph, and document databases, have come into the market, relational databases continue to be the most popular DBMSes.

Changing with the times

What accounts for the relational database’s remarkable staying power? And what’s next for what many CIOs and CTOs consider a strategic platform in their technology architectures?

The list of factors contributing to the relational database’s decades of success includes a well-understood model of tables, columns, and rows; standards-based programming with SQL, Java, and other languages; support for ACID (atomicity, consistency, isolation, and durability) transactions, which guarantee data consistency; and the ability to support both online transaction processing (OLTP) and business analytics.

Yet, I believe the real secret to the relational database’s longevity may be its ability to evolve to keep pace with the ever-changing nature of data, applications, workloads, and deployment alternatives—both in the cloud and on-premises. Hundreds of new capabilities have been built into Oracle Database since Oracle V2 was released. The advances include client/server support (1985), Real Application Clusters (2001), Database Cloud Service (2013), Oracle Autonomous Database (2018), and AI Vector Search (2023).

And while structured data has been the norm since the 1970s and still accounts for much of the most valuable business data, such as financial transactions and customer records, Oracle Database and other relational databases like it have morphed into multimodal databases that accommodate objects, spatial data, documents, graphs, and other unstructured data types.

Polyglot, cloud native developers

Here’s another way of thinking about the history of the relational database: 50 years is roughly equal to two generations of workers. So you might ask whether database technology used by our grandfathers will continue to appeal to a new generation of users, administrators, data scientists, and developers.

Given the relational database’s proven adaptability, I see no reason it shouldn’t continue to be a leading data management platform for many years to come. Look no further than Oracle’s native support for vector data and vector search in Oracle Database 23ai, as well as generative AI capabilities for its APEX and SQL Developer tools.

I think of these and other new capabilities, such as JSON Relational Duality views, which expose data in relational tables as JSON documents and vice versa, and support for retrieval augmented generation (RAG), which lets LLMs use a company’s data to improve response accuracy, as future proofing the database for customers and development teams. This investment in leading-edge research and development bodes well for the relational model in general and for Oracle Database in particular.

Developers now have a head start to build new applications and enhance existing ones with innovative 23ai features that simplify development of modern data-driven apps.”

Gerald Venzl Senior Director of product Management, Oracle

I think of these and other new capabilities, such as JSON Relational Duality views, which expose data in relational tables as JSON documents and vice versa, and support for retrieval augmented generation (RAG), which lets LLMs use a company’s data to improve response accuracy, as future proofing the database for customers and development teams. This investment in leading-edge research and development bodes well for the relational model in general and for Oracle Database in particular.

The big question is whether newly minted developers will embrace relational databases as they build new applications for mobile, AI, fintech, and other scenarios.

When I worked at Oracle a few years ago, we sometimes referred to this emerging constituency as “polyglot, cloud native” developers. This new generation of technologists uses multiple programming languages and tools, including cloud databases, to build and deploy applications. The Oracle Developer Resource Center is full of languages, SDKs, documentation, and code examples for this way of working.

All of this helps explain why Oracle took the unprecedented step of introducing Oracle Database 23c Free—Developer Release a full six months before the official launch of Database 23c, now 23ai, at Oracle CloudWorld in September 2023. Significantly, it was the first time Oracle has ever released a new version of its database to the developer community in advance of a general release.

“Developers now have a head start to build new applications and enhance existing ones with innovative 23c features that simplify development of modern data-driven apps,” explained Gerald Venzl, Oracle senior director of product management, in a blog post.

With Database 23c, Oracle introduced the idea of “declared intent” as a faster, better, and easier way to build applications compared with hand coding. The new technologies that enable this kind of automation include natural language commands, AI Vector Search, Oracle’s low-code APEX development framework, and JSON Relational Duality views. This revolutionary approach, Oracle says, represents “the future of data and application development.”

The multicloud future

The relational database has come a long way since those early days when Oracle V2 ran on a Digital Equipment Corporation PDP-11 minicomputer. Today, Oracle Database has become the system of record for hundreds of thousands of enterprises around the world.

Oracle Database runs on the latest computer hardware in private, public, hybrid, and multicloud environments, with advanced deployment options giving customers more choices and flexibility. For example, the Oracle Database@Azure service is now available in the Microsoft Azure East US Region. Oracle will operate and manage Oracle Exadata Database Service, the first of several planned Oracle Database services to run on Oracle Cloud Infrastructure (OCI), in Azure data centers.

The relational database has evolved from basic queries and simple joins in the 1970s to a secure, high performing platform supporting AI application development and much more in the 2020s. We can only imagine what the future holds for the relational model in another 50 years.

John Foley is editor of the Cloud Database Report and a vice president with Method Communications.

This article was updated on May 8 to account for an Oracle Database 23c name change to Oracle Database 23ai.


View more Oracle Connect articles

Oracle Chatbot
Disconnected