1. Using Oracle IT Analytics
  2. Maximize Performance and Availability
  3. Maximize Application Server Performance
  4. Analyze Application Server Workload Performance

Analyze Application Server Workload Performance

App Server Performance Analytics lets you analyze the workload performance of your application servers through measuring the volume of requests processed and the average response time.

Using App Server Performance Analytics, you can:

  • Identify the peak time with a high volume of server requests.

  • Analyze changes in workload performance brought about by changes in garbage collection overhead.

  • Identify how changes in JDBC connection time affect workload performance.

  • View information for the total request count and total request count time.

From the main page of the application, you can view workload performance data across your application servers, by:

  • Server Request Count: This is the total number of requests serviced by the application server

  • Server Request Processing Time: This is the time taken by your application server to process each request

You can drill down to view detailed information about the volume of processed requests and average response time through the following:

  • A distribution chart to view the percentage of servers with increasing, decreasing, or constant number of server requests and processing time.

  • A Distribution Analysis histogram that groups application servers according to percentage time in garbage collection and total JDBC connection waits. From this chart, you can drill down to view detailed information for the application servers within a selected bucket.

From the Workload Performance drill-down page, you can view a breakdown of the servers in the bucket that you had selected in the Distribution Analysis histogram. This information is presented in the form of a heap map, where each cell represents a server and the color of the servers is determined by the rate of change in the metric selected.

Note:

By default, information for the top five application servers is loaded, but you can choose to view information for all application servers in the selected bucket.

You can view the server breakdown according to the following:

  • Server Request Count

  • Server Request Processing Time

  • Garbage Collection Invocation Count

  • Percent Time in Garbage Collection

  • JDBC Data source Wait Count

  • JDBC Connection Request Count

  • Heap Usage After Garbage Collection

Use the drop-down list to select what metrics to group application servers by.

Select a single server, multiple servers, or a group of servers to view detailed information for the following:

  • Total Server Request Count: You can view the total number of server requests serviced by the selected application servers and the percentage change over the selected time period.

  • Average Server Request Processing Time: You can view the average time taken by the selected application servers to process requests, and the percentage change over the selected time period.

  • Total Garbage Connection Invocation Count: You can view the number of times that the selected servers called for garbage collection and the percentage change during the specified time period.

  • Total JDBC Datasource Wait Count: You can view the total JDBC data source waits across the selected application servers, and the percentage change during the specified time period.

  • Total JDBC Connection Request Count: You can view the total JDBC connection requests across the selected application servers, and the percentage change during the specified time period.

  • Total Heap Usage After Garbage Collection: You can view the total heap usage for the selected application servers after garbage collection is completed and the percentage change during the specified time period.

  • Server Trend: This chart displays the trend of the selected servers over the specified time period.

You can also perform an advanced analysis through the following:

  • Forecast Trend: You can view the forecasted trend of server performance in terms of garbage collection overheads or JDBC data source count.

  • Distribution Analysis: You can view the distribution of the percentage of time spent in garbage collection or the JDBC connection waits for each of the selected servers.

  • Correlation: You can view the correlation between the metric that most affects the Server Request Processing Time, according to specific measures. For example, you can view the degree to which the Percent Time Spent in Garbage Collection correlates with the Server Request Processing Time, by Server Request Count.