APM's capacity analysis report shows how many instances of your app are running and how busy they are. This information is useful to analyze whether your app has enough instances deployed to support request loads. You can then use this information to fine-tune your configuration for optimal performance.
The capacity analysis report is only available for Python and Ruby language agents.
View the capacity analysis report
To view the capacity analysis report for your app:
- Go to one.newrelic.com > APM > (select an app) > Reports > Capacity.
- Use any of New Relic's standard page functions to drill down into detailed information.
The report includes three charts.
Capacity chart | Description |
|---|---|
App instance busy | This chart shows how busy are your application's instances as the percentage of time to process requests. This includes the average, minimum, and maximum percentages for any time period on the chart. Use this information to determine if you have the right number of instances for your application. As application utilization approaches 100%, your application needs more instances to handle incoming requests. |
App instance restarts by host | This chart shows the number of instance restarts for each time interval. Use this information to determine if your application instances are restarting too frequently. To isolate activity to a specific host, select other hosts' titles on the legend below the chart to hide or return them to the chart. OR: Select your choice from the Hosts list on the APM Summary page. |
App instance analysis | This chart shows the total number of instances running, along with the concurrent instance load. The concurrent instance load is the number of fully busy instances that would be needed to handle the load on the app. New Relic bases this on the average app instances value from the App instance busy chart, multiplied by the number of instances you have running. |
Python single or multi-threaded apps
For the Python agent, capacity analysis reporting only works for traditional single or multi-threaded applications. This is because the way New Relic calculates a system load does not translate to co-routine based systems.
When using a traditional single or multi-threaded application, the metric information generated by the Python agent is captured and reported only if the optional C extension component can be compiled and installed. When you use co-routine based systems such as gevent or eventlet in conjunction with a WSGI server, the capacity analysis report will not show any data except for agent restarts.
Analyze your data
APM includes several reports in the user interface. To gather, analyze, and visualize data about your software in other formats, use query builder.