Big Data and NoSQL are hot topics and are considered by many to be the ultimate new solution for handling an ever-growing flood of data in today’s world.
The ideas behind NoSQL and non-relational concepts aren’t all new. A look back in history shows us that some past approaches were very similar to those of today. For example, Software AG’s Adabas (Adaptable Database System), introduced in 1969, exemplified the technologies and concepts developed in the pre-relational era to efficiently represent complex transaction and data model logic in databases, without the need for normalization (i.e. putting everything into tables, rows and columns). Even today, this focus is unique to Adabas, which can process more than one million database operations per second.
Adabas, as one of the ancestors of NoSQL concepts were obviously ahead of their time. If we look at the first 30 years of database development starting in 1960, we see a transformation from specialized non-normalized data processing systems to normalized and relational systems. The success of relational database management systems (RDBMS) and the SQL data query language had a strong influence on the normalization trend between 1990 and early 2000, as did the development of relational-based data warehouse (DWH) systems.
The normalized world really began to change in early 2000 with the development of Big Data and NoSQL technologies. IT again started moving from a normalized database world toward highly specialized data processing systems.
We are now at a stage to combine the power of specialized database systems that enables a new era of data architectures and innovative data-driven solutions.
Today’s customer projects aim to extend the transactional data processing capabilities of Adabas with analytical big data systems (e.g. Hadoop, Spark, Apache Cassandra, ElasticSearch or Software AG’s market-leading streaming analytics Apama and in-memory data platform Terracotta) in order to gain more insights into data and to drive new digital business initiatives (e.g. mobile banking, citizen online services, marketing initiatives, fraud detection and auditing).
This combination can easily be achieved by mapping the Adabas data model to key-value or document-oriented data as well as column and graph-oriented data structures. And with the connectivity of Adabas the data can be easily accessed and replicated in real-time to Big Data systems using standard interfaces and protocols (e.g. REST API). A real win-win situation.
As part of Software AG’s Adabas & Natural 2050+ initiative we continue to further improve these Adabas capabilities – so stay tuned what the future will bring.