Dynamic Cache Statistics Collector and Visualizer for IBM WebSphere Application Server

1. Can you provide a sample use of the DynacacheStatistics.py script?

$WAS_HOME\bin> wsadmin -f DynaCacheStatistics server1 server1_stats.txt "-nodeName node1 
	-cacheInstance services/cache/cacheone -sleepUnit minutes -sleepInterval 15 -fileAppend"

Note: The command constituting the above two lines should be run in a single line on the wsadmin prompt.

Back to top

2. How do I export cache statistics collected to a CSV file that can be imported into Microsoft Excel?

Set the outPutFileExt property in setup.txt to csv and the quantitativeLevelUpperBound property to 0, and then rerun the DynaCacheStatisticsParser.bat file. This action will generate a CSV file of the collected statistics that can be imported and charted in Excel.

Back to top

3. How do I prevent the normalization of the cache statistic values?

Set the quantitativeLevelUpperBound property in setup.txt to 0 and rerun the DynaCacheStatisticsParser.bat file.

Back to top

4. How do I convert the SVG file generated to a JPEG file?

Use the Apache Batik open-source SVG rasterizer, as follows:

java -jar batik-rasterizer.jar -m image/jpeg -q 0.8 chart.svg

(See this article on developerWorks: Tip: Create JPEGs automatically with SVG: Use Scalable Vector Graphics to create images by the dozen.)

Back to top

5. Where can I find further information about the DynaCache component of WebSphere®?

• Enhancing the performance of WebSphere Commerce applications through dynamic caching
• Tutorial: Improve WebSphere Commerce performance with dynamic caching
• Caching WebSphere Commerce pages with the WebSphere Application Server dynamic cache service (white paper)
• Mastering DynaCache in WebSphere Commerce (redbook)
• WebSphere Dynamic Cache: Improving J2EE™ application performance (article in the IBM Systems Journal)
• Dynamic cache and data replication service tuning guide

Back to top

6. Where can I find more information about the DynaCache Mbean?

Please see the Disk cache enhancements page.

Back to top

7. What are some of the known limitations of this technology?

The cache statistic visualization is static. The technology cannot collect and visualize the statistic at the same time. It cannot show the running state of the system.

Back to top

8. Does this tool work with Sun's 1.5 Java™ Virtual Machine (JVM)?

Yes, but you must add a xerces.jar in the lib directory.of the unzipped dynacache.zip. The xercesImpl.jar file can be downloaded from the Apache Web site.

Back to top

9. Do you have some data that we can use for testing the Visualizer?

Sample data files (trade1.txt and trade2.txt) are included in the downloadable dynacache.zip file.

Back to top

10. What kind of performance overhead does the tool have?

The tool has two aspects:

• Data collection: The statistics collector script that runs on the production servers can be set up to run at different intervals. Therefore, it can collect statistics every 4 hours or every 4 minutes. The degree of intrusiveness is controlled by you.
• Data visualization: Remember that the visualizer is a stand-alone tool, so, from a production server point of view, it has no impact if this tool is run on a different server after the data has been collected.

The overhead of invoking JMX™ Mbean API calls is unintrusive as compared to the monitoring provided by services such as PMI or third-party JMX vendors.

Back to top