Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.odigos.io/llms.txt

Use this file to discover all available pages before exploring further.

See it in action!

The clip below shows the full path on a host: enabling profiling.enabled via odictl, adding a Pyroscope destination, and the resulting flame graph in the Pyroscope UI. End-to-end VM Agent profiling demo: enable profiling, add destination, view flame graph

What is continuous CPU profiling?

CPU profiling samples what your processes are doing on the CPU. Over time, those samples form a flame graph that shows which functions and libraries consume the most CPU time. Unlike traces, which tell you that a request was slow, profiles show where CPU time was spent. Use profiles after traces (or metrics) point you at a hot service.

How it works on the VM Agent

  1. You enable the profiling pipeline in agent config (profiling.enabled: true).
  2. odigos-otelcol runs the OpenTelemetry eBPF profiler receiver on the host.
  3. Profiles are exported as using OTLP Profiles protocol to a destination which supports OTLP profiles.

Prerequisites

Getting Help

If you have any issues, or require our assistance, please open an issue in GitHub, or reach out to us in Odigos Slack