Connection Pooling and Memory table question

Jun 23, 2011 at 9:58 PM

Hi, I was wondering how the connection pooling in effiprozdb worked. I ask because I am expirencing some performance issues using microsoft TPL for executing db calls. I have auto shutdown set to false and auto commit set to true. I know this database gets good performance if a single connection is used but since I spawn off a new thread with every db call, I close all my connections after the db call and create new ones with new calls. Im testing performace by inserting 5000 rows and it's very slow. Could this be because there isn't enough connections to handle the load?

Also my tables are of type MEMORY and have a huge .script file with a bunch of inserts that never get persisted into a data file. Is this because of the table being of type memory? If so how do I create a table of type file instead? I would like to change whis since it's taking way too much ram. Here is a sample of my connection string

"Connection Type=File ;auto commit= true; auto shutdown= false; user=sa;Password=;initial catalog=C:\\Users\\jay.LP\\Documents\\LPTestDB;"

 

Jun 24, 2011 at 12:36 AM

Never mind. Didn't think this was a memory only database. You should write that on your web site so people don't get confused.

For anyone that wants to know more about this here is the link.

http://hsqldb.org/doc/guide/ch01.html#N1023C