1. Home
  2. Open-Source Database Software
  3. PostgreSQL Reviews
  4. User Review of PostgreSQL
PostgreSQL - the last database you'll ever need
August 02, 2016

PostgreSQL - the last database you'll ever need

David McCann | TrustRadius Reviewer
Score 10 out of 10
Vetted Review
Verified User

Overall Satisfaction with PostgreSQL

PostgreSQL serves as the database for our web-based product. As such it is used by the entire organization and is an integral part of the product.
  • Broad array of custom functionality/formulae built atop standard SQL statements
  • Transactional support for schema migrations
  • Great low-level performance tuning capabilities
  • Performant implementation of bulk merge (upsert)
  • Some of the custom SQL functions thatPostgreSQL provides could be optimized. Specifically I've seen that the "is contained within" operator for inet columns is slow to the point of being unusable in bulk (e.g. as a join criteria).
  • By pushing a lot of logic into database views and triggers, PostgreSQL provides the ability to write much slimmer and more maintainable controller code, reducing development costs.
  • PostgreSQL's upsert capabilities enabled us to go to scale much more easily and quickly, allowing us to increase our number of customers.
  1. MySQL - doesn't have transactional support for schema migrations, and has a more restrictive license
  2. DB2 - not free
  3. SQLite - doesn't have full foreign key constraint support
  4. CouchDB - slightly different use case, however PostgreSQL does offer performant to_json functionality that overlaps with this attractive CouchDB capability
Overall, PostgreSQL is the best default choice for any database-driven web application with a fairly common set of requirements.
Visual Studio.NET, TeamCity, Apache Web Server
PostgreSQL is in my opinion the best open-source option for any enterprise-level product requiring a standard relational database. Areas where it might be considered not ideal would include: smaller-scale projects requiring a database (MySQL might be an alternative here), or document stores with unstructured data (some would argue that NoSQL options are better here).