Efficient Oracle Database Usage By Java and J2EE Applications

In case you haven't guessed, I like producing presentations. In the next two weeks or so, I intend to produce a new presentation on the the "Efficient Use of Oracle By J2EE and Java Applications". To whet the appetite this will include:-

1. Connection management best practices:-

• Connection pool usage
• Setting the connection pool min and max settings to be the same to avoid connection storms
• Oracle 11g server side connection pooling
• Using the 11g to throttle connection rates

2. Efficient cursor management

• The statement Vs preparedStatement API.
• Avoiding hard parsing
• Avoiding soft parsing
• Parameters that affect parsing: session_cached_cursors,
• cursor_space_for_time and cursor_sharing

3. Minimizing network round trips and leveraging bulk operations

• The JDBC fetch size
• Statement batching in JDBC
• The Oracle array interface
• Leveraging Oracle bulk operations
• Intelligent Oracle statement feature usage, specifically usage of: merge
• statements, in line views, in line PL/SQL functions, sub query factoring.
• The client side result cache
• Programming and design styles for minimizing application 'Chatter'
• Caching

4. Design and coding according to the putting the processing close to resource that requires it, specifically:-

• Do not be precious about Java or the database; it is not necessarily a good
• thing to do everything in the database.
• Do not re-invent database functionality in Java

5. Best practices for logger buffer usage

• JDBC auto commit.
• Batch and asynchronous commits as introduced in Oracle 10g.

Watch this space . . .