This blog promotes knowledge sharing through experience and collaboration. For more product information, visit our WebSphere Commerce CSE page. For easier navigation, utilize the Categories to find posts that match your interest.
Still Using Javacores for Performance Analysis?
But reading Javacores can be time consuming. You not only need to compare threads from the same Javacore, but also compare activity across multiple Javacores. That can take a while.
With the recommended options, the Health Center Agent collects thread data every 30 seconds, 24x7, and saves it into .hcd files every 15 minutes. You can then grab the .hcd file and load it into the WebSphere Commerce Health Center Report tool to obtain a report.
The report includes thread data presented at three different abstraction levels. This allows you go from a high level overview, which compares multiple threaddump times, to the thread/stack level as you would have in a Javacore.
The WebContainer Base Activity chart gives you a high level overview of the top activities on each threadump, and makes it easy to compare activities across threaddump times. To present a summary view, the chart uses the base activity, which is the activity at the top-most section of the stack:
Next you can review the WebContainer Detailed Activity table, which will show the complete activity chain for each thread and threaddump time. This, for example, allows you to see if the base activities (e.g. waiting for DB2) have a common root:
And finally, by clicking a thread/time, you can access the Thread Details pop-up, which will give you access to the native Javacore-like stack:
Besides using the charts and tables on the web, you can download a Zip containing the thread data in Javacore-like files, which you can use for further analysis or for reviewing non-WebContainer threads:
Javacores continue to be a critical tool for troubleshooting, but for the specific case of thread analysis, the WebSphere Commerce Health Center Report tool can not only help you save time, but also give you insight and recommendations for your configuration and runtime performance. The tool automates thread data collection, and with samples captured every 30 seconds, it helps you ensure you don't miss captures during events that need to be investigated. Check the documentation for instructions on how to set it up and to access the demo reports.