Discussing Converged database versus Specialized database engines

Since I started working with databases there have  been multiple paradigm shifts in management of data.

Some notable changes include

> Moving to different data types , e.g image , video , clob, blob, xml, cursor, table, varbinary, binary , spatial types geometry , spatial geography types

> Number of users & uses  , particuarly through the Internet explosion where the application requirements rapidly multiplied . This new class of applications required storage and search capability for a wider range of entities

> The growth of the Developer community is in parallel to the  rise of special purpose database engines . There is an increasing emphasis on  Rapid development , leading to process changes and flexibility in the way applications are architected and deployed.

Developers require a more efficient method of storing data   either schema or schema-less.   Schema-less data , such as document storage , and mapping the schema to the  application data object , deserves some attention - specifically around  data integrity rules  and where to maintain the rules. Traditionally the advice is to maintain the integrity rules as close to the data as possible - so the rules move with the data.   Now , we're witnessing a distributed model with data across multiple data stores\engines and rules in various phases of the transaction.

Cross-roads in the industry

In general terms , we're at a cross roads in the industry . On the one side , there is the rise of Agile development on such platforms as AWS  - with   package automation around multiple DB engines - available quickly , without the overhead of infrastructure  management  and rapid access to different DBMS engines with layered storage costs. 

On the other side - there are the CTOs and DBAs pushing for a converged database service. This is  noticeable in environments with thousands of databases , supporting a wide range of DIY & vendor applications

The justification for the converged approach is to decrease the fragmentation of data , rationalizing data schemas , challenging the need for separate database engines. The DB engine Vendors are responding by adding new features and data type support with every version upgrade . Most of the large DBMS vendors such as SQL Server, Oracle, PostgreSQL have support for JSON. 

Some other related topics , but contribute to a broader conversation

1) Data Lake -  A data lake is a storage repository holding large amounts of raw data in its native format until it is needed for analytics applications. Hadoop is a good example accompanied with storing data on cheaper storage HDFS .

2) BlockChain technologies fit into these developments. Consider the latest announcements from Microsoft  - SQL Server Ledger Tables and  Blockchain Oracle