Oracle Autonomous JSON Database
On the 13th of August 2020, Oracle announced the general availability of a “new” public cloud service: the Oracle Autonomous JSON Database.
Not so new indeed as it leverages the existing Oracle Autonomous Database platform running already the Autonomous Data Warehouse (aka ADW) and Autonomous Transaction Processing (aka ATP).
The idea behind this new service is to make available the Autonomous Database capabilities (multi-model support, zero administration, independently scale-out compute and scale-out storage infrastructure, 99.995% availability, self-patching, automatic-scaling, state-of-the-art security…) to the Web/Frontend/Backend/Full-Stack… developers while providing new innovations in the world of storing and processing hierarchical data, here JSON data.
In its launch announcement post, Beda Hammerschmidt highlighted some of the differentiators with MongoDB such as unrestricted ACID transactions, partial updates support, 2x Speed-up, simple API (known as Simple Oracle Document API or SODA), Low-Code platform (known as APEX) that allows developing web applications (mobile, responsive…) very quickly, expose data through web services or REST API natively…
All of that cheaper than MongoDB Atlas:
Another cool thing for developers is that AJD is available through the Oracle Cloud Always Free Tier for free forever (choose ATP).
Finally, Multi-Model support means developers can work on Hierarchical data (JSON) and Relational data at the very same time and benefit from the advantages of both delivering true freedom:
- Ingest JSON documents and use advanced SQL capabilities such as MATCH_RECOGNIZE function, Analytic functions, Materialized views, Row Level security…
- Expose JSON documents build from Relational data through REST APIs (see Oracle REST Data Services natively available with Autonomous Databases) using the JSON_* SQL functions
- Perform advanced Full-Text searches on JSON data using fuzzy or wildcards… operators
- Build Machine Learning models right on top of your JSON documents and score new ones
- Join distinct collections between them, or with Relational data, Spatial data, XML data…
My best advice: try it!!
