#tonic #tracing #propagation #fastrace #traceparent #distributed-tracing #client-server

fastrace-tonic

A tonic instrument for propagating trace context for fastrace

1 unstable release

new 0.1.0 Mar 18, 2025

#669 in Debugging

Apache-2.0

10KB
70 lines

fastrace-tonic

Crates.io Documentation MSRV 1.75.0 CI Status License

fastrace-tonic is a middleware library that connects fastrace, a distributed tracing library, with tonic, a gRPC framework for Rust. This integration enables seamless trace context propagation across microservice boundaries in gRPC-based applications.

What is Context Propagation?

Context propagation is a fundamental concept in distributed tracing that enables the correlation of operations spanning multiple services. When a request moves from one service to another, trace context information needs to be passed along, ensuring that all operations are recorded as part of the same trace.

fastrace-tonic implements the W3C Trace Context standard for propagating trace information between services. This ensures compatibility with other tracing systems that follow the same standard.

Features

  • 🔄 Automatic Context Propagation: Automatically inject trace context into outgoing gRPC requests.
  • 🌉 Seamless Integration: Works seamlessly with the fastrace library for complete distributed tracing.
  • 📊 Full Compatibility: Works with fastrace's collection and reporting capabilities.

Installation

Add fastrace-tonic to your Cargo.toml:

[dependencies]
fastrace = "0.7"
fastrace-tonic = "0.1"

Server Integration

Apply the FastraceServerLayer to your tonic server:

use fastrace_tonic::FastraceServerLayer;
use tonic::transport::Server;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Initialize fastrace reporter.
    fastrace::set_reporter(ConsoleReporter, Config::default());
    
    // Add FastraceServerLayer to your tonic server.
    Server::builder()
        .layer(FastraceServerLayer)
        .add_service(YourServiceServer::new(YourService::default()))
        .serve("[::1]:50051".parse()?)
        .await?;
    
    fastrace::flush();

    Ok(())
}

Client Integration

Apply the FastraceClientLayer to your tonic client:

use fastrace_tonic::FastraceClientLayer;
use tower::ServiceBuilder;

async fn make_client_call() -> Result<(), Box<dyn std::error::Error>> {
    // Create channel with FastraceClientLayer.
    let channel = tonic::transport::Channel::from_static("http://[::1]:50051")
        .connect()
        .await?;
        
    let channel = ServiceBuilder::new()
        .layer(FastraceClientLayer)
        .service(channel);
        
    // Create client with the enhanced channel.
    let mut client = YourServiceClient::new(channel);
    
    // Make calls as usual.
    let response = client.your_method(Request::new(YourRequest {})).await?;
    
    Ok(())
}

Example

Check out the examples directory for a complete ping/pong service example that demonstrates both client and server tracing.

To run the example:

  1. Navigate to the example directory:

    cd example
    
  2. Start the server:

    cargo run --bin server
    
  3. In another terminal, run the client:

    cargo run --bin client
    

Both applications will output trace information showing the request flow, including the propagated context.

How It Works

  1. When a client makes a request, FastraceClientLayer detects if there's an active trace and adds a traceparent HTTP header with the trace context.
  2. When a server receives the request, FastraceServerLayer extracts the trace context from the traceparent header and creates a new span as a child of the received context.
  3. If no trace context is provided, the server creates a new root span.

This process ensures that all operations across services are properly connected in the resulting trace, providing visibility into the entire request lifecycle.

License

This project is licensed under the Apache-2.0 license.

Dependencies

~2.7–8.5MB
~66K SLoC