You can now follow a request through every Python service it touches and see where it slowed down or failed. Distributed tracing in Atatus Python APM records the timing and errors at each step as a request moves across microservices, cloud environments, containers and service meshes.
The agent collects spans automatically from the major Python frameworks and libraries, so you can find bottlenecks and failures without writing tracing code.
What you can do
- Trace end to end: Follow each request across services and components, span by span.
- Pick a trace view: Read a trace as a tree chart, a flame graph or a span list.
- See errors in context: Errors and exceptions appear inside the trace where they happened, with stack traces and context.
- Track key metrics: Follow response times, request rates and error rates for your services.
- Control data volume: Configure sampling to manage how much trace data you send while still capturing the traces that matter.
- Filter with tags: Add custom tags and annotations to traces to filter and analyze them.