Drill down on graphs

Webmetrics graphs are interactive: you can mouse over graph elements to display information about the hourly average, number of errors, sample time, and so forth. In some of these graphs, you can also click on data points to drill down on that information and get even more detail of performance for that time period.

The following graph types allow you to drill down on performance information:

  • For Application Monitoring services:
    • Complete Transaction over Time
    • Load Time - City by City (Scatter)
  • For Website Monitoring services:
    • Load time by Period
    • Load Time - City by City (Scatter)
  • For Web Service and Web Service Transaction monitoring services:
    • Response Time by Period
    • Response Time - City by City (Scatter)

An example:

  1. A user receives a weekly report via email and looks at the performance of a service over the course of the week. In this example, the Sample_Transaction service is an Application Monitoring FullPage Gold with a two-step scripted transaction.
  2. The user is interested in viewing the Complete Transaction over Time graph and clicks on it, which loads a Flash-based interactive graph for that time period. Mousing over data points on the graph shows the value at each point.
  3. The user is interested in viewing performance data for Wednesday, so she clicks on that average in the graph, which then opens a new window, graphing the performance for Wednesday on an hourly basis. Like the weekly performance graph, this graph is interactive and allows her to mouse over data points to see the actual value that was graphed.
  4. The user then chooses to view the performance for 8:00am, so she clicks on the average for the 8:00 hour. This generates a scatter plot of all the samples taken during that hour. Here, since the user has enabled baseline agents, only samples taken from those agents are displayed.
    • Mousing over the samples at this level displays information about the agent that took the sample, the time it was taken, and the load time of the sample.
  5. The user clicks on a sample in the scatter plot to bring up the raw log file of all the samples taken for that service from that agent on that day.
  6. Finally, the user can click on the fullpage breakdown icon and can view detailed information for each item in the sample, potentially helping to identify where any performance issues lie.

See also