I have to do a bulk data upload using JMS deploy in a glassfish 2.1, to process and validate data in a Oracle 10g DB System before it is insert.
I have my web interface that loads a file and then delegates the process to a Stateless Session Bean which read a N number of lines and after that send a message to a JMS Queue. The JMS has to parse the each line, validate them with the data already in the DB, and finally persist the new data.
This process is high process consuming, and I need to improve the performance time. I tried to change glassfish default JMS and JDBC pool size, but I have no a huge difference.