>Home>Learn XQuery>XML eXtra Archive>August, 2005 Print

DataDirect XQuery® Beta - 1 month remaining!

Welcome!

Over the last few months, I have had the opportunity to talk with many of you about XQuery, XQJ, DataDirect XQuery®, and your projects. I have been very impressed by the quality and experience of the people who have signed up for our beta program. I was also very surprised by the large number of people who signed up!

There is one month left in the beta program and we are still interested in learning about how you are using DataDirect XQuery® in your products and architectures. We are glad to offer advice to help you. Please contact the beta coordinator to set up an appointment. And if you have a technical question, send an email to our tech support team.

-- Nancy Vodicka, XML Program Manager

August 2005: Table of Contents

1. From Our Beta Participants
2. Querying from the GUI: Performance and Memory Issues

1. From Our Beta Participants

This month's beta spotlight is on a project by a leading ERP vendor at an Australian chemical manufacturer. The challenge is to integrate a new ERP system (DB2 database) with an Advanced Planning System (a combination of XML, flat file and OODB with SQL Server database considered). XQuery plays dual roles in this project - both in the integration middleware piece and also for reporting data from the databases. Paul Keating, Project Manager, says "I prefer XQuery to XSL, it's easier and the differentials between XSL and XQuery are huge. XQuery saves development time."

2. Querying from the GUI: Performance and Memory Issues

1. Never measure the performance of DataDirect XQuery® in or Stylus Studio. Inside these GUIs, performance is *much* slower than in a Java program, since they spend a good deal of time updating the display. Also, these GUIs initialize each query before running, which requires classes to be loaded, schema information to be loaded, and queries to be prepared - resulting in very different timings than you will have in a program that initializes only once.
2. If you are processing large XML files in your query and performance is slow, try increasing your heap size - this can dramatically improve your performance.

You can set the initial heap size and maximum heap size for the Java Virtual Machine on the command line. On Sun's JVM, use the following options:

• initial heap size: By default, this is the larger of 1/64th of the machine's physical memory on the machine or some reasonable minimum. You can override this default using the -Xms command-line option.
• maximum heap size: By default, this is the smaller of 1/4th of the physical memory or 1 GB. Before J2SE 5.0, the default maximum heap size was 64 MB. You can override this default using the -Xmx command-line option.

If you are running your query in or Stylus Studio, you can set the heap size in the tool.

• Eclipse uses 64 MB by default. You can set it to 256 MB using the following command line:
  eclipse.exe -vmargs -Xmx256m
• In Stylus Studio you can control the heap size through the Java Virtual Machine page in
  Tools | Options.