Re: [JDBC] [HACKERS] Plan invalidation vs. unnamed prepared statements

2007-03-08 Thread Тимчишин Виталий
I am from pgsql-jdbc, so I may not be "in the thread", so please ignore places where my misunderstanding goes out. The main two questions, IMHO, is: 1) What is the key to plan cache. Current option is some statement key (id). Another option would be statement text (you still need to store it

Re: [JDBC] [HACKERS] Plan invalidation vs. unnamed prepared statements

2007-03-06 Thread Dave Cramer
I think C is how the JDBC driver is written. We name the statements if they have been used more than prepareThreshold times. So we have a mechanism by which to allow statements to be cached, or not. Dave On 6-Mar-07, at 1:14 PM, Tom Lane wrote: Gregory Stark <[EMAIL PROTECTED]> writes: