Profiling resultSet.getXXX methods

Hi,

I'm working on a JDBC performance impovement project. Researching JDBC best practices it's my understanding that you should avoid getObject behavior and use type specific methods (getInt, getFloat, etc) I'm finding that in some instances that performance is worst when I use the type specific methods. Also I find that the performance results are inconsistent with either getObject or the type specific methods. It's very interesting that sometimes it takes 10 miliseconds while others are 0 milliseconds. I'm wondering if there is any signifigances to the 10 milliseconds because either I see 0 milliseconds or 10 milliseconds.

Any suggestions would be appreciated...

[696 byte] By [bob000000a] at [2007-11-27 5:00:52]