Overview

The profiling flame graph visualization represents a breakdown of top consuming lines of code such as CPU and Memory. Add this widget to visualize stack traces of your profiled applications and accurately identify frequent resource requests.

Setup

Configuration

1. Scope your profiling data with tags. For example, host, container_name, service, env, or version.
2. To select the resource click the dropdown menu next to Show. Options can include CPU Time, Allocated Memory, or Thrown Exceptions.
3. Click the dropdown menu next to by and for to select the frame granularity and code provenance, respectively.
4. Give your graph a title or leave the box blank for the suggested title.
5. Click Save.

Options

Advanced options and filtering

Click the three dot ellipsis to open Advanced options to specify coloring and resolution.

Customize your flame graph. Add graphing actions or filters in the Filter flame graph field.

Scope to endpoints

Filter on a specific endpoint, for total consumption (per Minute by Endpoint) or per request (per Endpoint Call).

Scope to functions

Filter on other criteria such as Method, Package, Thread name or Trace Operation.

Global time

Choose whether your widget has a custom timeframe or the dashboard’s global timeframe.

API

This widget can be used with the Dashboards API. See the widget JSON schema definition.

Further Reading

Más enlaces, artículos y documentación útiles:

Learn about Profile visualizationsDOCUMENTATION Building Dashboards using JSONDOCUMENTATION