One of my lessons-learned from JavaOne is the importance of choosing the right data access. I try to preach using the right tool for the right job and didn't realize that in my mind I had made Hibernate the choice for all data access.

Two talks by Cameron Purdy helped me expose my inner-hypocrisy.

• TS-1402,"Essential Lessons of Distributed Caching"
  • Details | Slides
• TS-5397,"The Top 10 Ways to Botch an Enterprise Java™ Technology-Based Application"
  • Details | Slides

The slides require a username and password - User Name: contentbuilder and Password: doc789

Each method of data access has its own strengths and weaknesses. ORM (e.g. Hibernate) is great for rapid development, large teams, data caching and I think best for most general uses. Whereas, JDBC is great for large reporting sets and direct access. The choice isn't always obvious and you could even use both.

A great reminder for me to always diagnose the problem before applying the solution. Thanks, Cameron!