Question 1

How to trace async functions in Rust with tracing?

Accepted Answer

Use the #[instrument] attribute on async fns or manually wrap futures with Future::instrument, as shown in the README. Avoid using span guards directly in async blocks, as they can cause incorrect span lifetimes.

Question 2

Tracing vs log crate: which should I use?

Accepted Answer

Use tracing for structured, context-aware diagnostics with spans and async support, especially in libraries or complex apps. Stick with log for simple, unstructured logging where you don't need tracing's advanced features.

Question 3

What is the best subscriber for tracing in production?

Accepted Answer

It depends on your needs: tracing-subscriber's fmt module is good for development, while tracing-opentelemetry integrates with OpenTelemetry for distributed tracing. Check the ecosystem crates for backends like Honeycomb or Loki.

Question 4

Can tracing be used without Tokio runtime?

Accepted Answer

Yes, the README specifies that tracing is maintained by the Tokio project but does not require the Tokio runtime, making it suitable for non-Tokio async runtimes or synchronous code.

Question 5

How to filter tracing events by level?

Accepted Answer

Use the with_max_level method when building a subscriber, as shown in the README example with tracing_subscriber::fmt().with_max_level(Level::TRACE). This controls which spans and events are collected.

Question 6

Is tracing suitable for high-performance applications?

Accepted Answer

It can be, but with caveats: the instrumentation overhead might impact latency-critical paths. Consider using level filtering or disabling tracing in production for hot code paths to mitigate performance hits.

tracing

What is tracing?

Overview

Use Cases

Best For

Related Projects

Found a gem we're missing?

Not Ideal For

Pros & Cons

Pros

Cons

Frequently Asked Questions