Background Fundamental Software is currently investigating and prototyping a fully cloud based web portal for the FPM application. It will initially provide a portfolio analysis view for clients and at a later phase, additional functions such as the import and edit of instrument and portfolio data will be possible.

Research & Development of SQL vs NoSQL The development team was tasked with investigating technologies with which to create the client and server sides of the FPM web portal. We explored both open source and Microsoft options.

In groups of two to three, each group researched different front or back end frameworks, and then presented their findings and some prototypes to the rest of the team.

Back End We are looking for a platform independent, open source database. Specific options chosen to look at more closely were MongoDB and PostgreSQL. Basically we need to decide whether to go with SQL or NoSQL..

This blog post tells the story of a project that started off using Mongo. As things went ahead, the team discovered that it was not the right choice because although their data initially did not appear to be relational, it was. They then moved on to MySQL.

If we want a flexible cache for massive volumes of unstructured data before displaying it or returning to the FPM application database for processing, then Mongo would present a fast, scalable solution.

Should we want to store structured, relational data and perform any complex queries, and if data consistency is more important than speed, PostgreSQL would be the better option.

What are your thoughts on or experiences with SQL vs NoSQL and cases for use of each?