The Amazing Babelfish for PostgreSQL by AWS

Some interesting news in the  AWS re:Invent 2020 - Keynote with Andy Jassy . 

Introduction of the new Babelfish for PostgreSQL , aiming to to offer a new translation capability to run SQL Server apps on Aurora PostgreSQL. One of the main selling points is the capacity to understand SQL Server proprietary dialect t-sql and the Tabular Data Stream (TDS) Protocol

Some underlying logic , which has led to the development & introduction of Babelfish for PostgreSQL:

1) Databases are  hard to manage   which partly has led to the massive uptake on managed RDBMS service. Despite growth in AWS managed DB services , the majority of DBs are still on-prem. A vast majority are hosted on SQL Server & Oracle . 

2) There is a large push on applications being supported\moving to MySQL & Postgresql

3) Amazon Aurora - 100% compatible as PostgreSQL\ MySQL. There has been a large uptake of Amazon Aurora Serverless - primarily for Dev environments

5) No question : Migration is painful - how can you move the application code ? AWS offering  Babelfish for Amazon Services

The devil is in the detail - when it comes to migrating heavily embedded t-sql based applications , where business logic can be spread across the database and the application, depending on the development methodology. 

I've previously been involved in these migration exercises - for example Oracle to DB2 LUW . Large amounts of resource were required . With multiple planned  application migrations aborted due to cost . 

Before embarking on this journey - some questions should be asked:

1) Is the end goal to migrate onto PostgreSQL? Will PostgreSQL satisfy the application requirements? Will the feature set on PostgreSQL fulfil the application requirements?

2) If it's a 3rd party product - will the vendor certify on PostgreSQL?

3) Is it more efficient\cost effective to build from scratch and migrate module by module?

An image of how AWS Babelfish for PostgreSQL , courtesy of Amazon