All Products
Search

This Product

    Managed Service for OpenTelemetry:Report Rust application data through SkyWalking

    Document Center

    Managed Service for OpenTelemetry:Report Rust application data through SkyWalking

    Last Updated:Mar 10, 2026

    When you run Rust microservices, you need distributed tracing to diagnose latency and track requests across services. The skywalking-rust crate is the official Rust agent for Apache SkyWalking, an open-source Application Performance Monitoring (APM) and distributed tracing system originally developed in China. SkyWalking is designed for microservices applications, cloud-native applications, and containerized applications in Docker, Kubernetes, and Mesos. It lets you manually instrument Rust applications and report trace data to Managed Service for OpenTelemetry over gRPC.

    For a complete working example that instruments an HTTP server built on the hyper framework, see the skywalking-demo repository on GitHub.

    SkyWalking span types

    SkyWalking requires manual instrumentation for Rust applications. You create spans in your code to represent units of work, and the agent reports those spans to the collector over gRPC.

    Span typeRoleExample
    Entry span (EntrySpan)Inbound request on the server sideAn HTTP server handler receiving a request
    Local span (LocalSpan)In-process work that does not cross a network boundaryA local function call or computation
    Exit span (ExitSpan)Outbound call from the client side; the client uses an exit span to inject trace context into the outgoing requestAn HTTP client call to a downstream service

    Instrument a Rust application

    Prerequisites

    • Protobuf is installed (unless you use the vendored feature; see Step 1).

      macOS

      brew install protobuf

      Debian/Ubuntu

      sudo apt install protobuf-compiler

    • To obtain an endpoint of SkyWalking, perform the following steps:

      1. Log on to the Managed Service for OpenTelemetry console.

      2. In the left-side navigation pane, click Cluster Configurations. Then, click the Access point information tab.

      3. In the top navigation bar, select a region. In the Cluster Information section, turn on Show Token.

      4. In the Client section, click SkyWalking.

      5. Copy the endpoint from the Related Information column.

      SkyWalking endpoint information

      Note

      If your application runs in an Alibaba Cloud production environment, use a Virtual Private Cloud (VPC) endpoint. Otherwise, use a public endpoint.

    Step 1: Add the SkyWalking dependency

    Note

    The examples in this topic use SkyWalking 0.8.0.

    Choose one of the following methods:

    Option A -- Edit Cargo.toml directly (recommended)

    Add the following line under [dependencies]:

    skywalking = { version = "0.8.0", features = ["vendored"] }

    The vendored feature compiles protoc from source during the build, so you do not need to install protoc separately.

    Option B -- Use the cargo add command

    cargo add skywalking --features vendored

    Step 2: Create spans to instrument your code

    Import the SkyWalking modules and use entry, local, and exit spans to trace requests across your application.

    use skywalking::{reporter::grpc::GrpcReporter, trace::tracer::Tracer};

    Server-side instrumentation (entry span)

    Create an entry span for each inbound request your application handles. The server uses an entry span to obtain the tracing context from an HTTP request.

    let mut ctx = tracer.create_trace_context();
{
    let _span = ctx.create_entry_span("handle_request");

    // Your request handling logic goes here.
}

    Client-side instrumentation (exit span)

    Create an exit span before calling a downstream service. The client uses an exit span to inject the tracing context into the outgoing request.

    let mut ctx = tracer.create_trace_context();
{
    let _span = ctx.create_exit_span("call_downstream", "downstream-service:8080");

    // Your outbound call logic goes here.
}

    In-process instrumentation (local span)

    You can use a local span to instrument an application in the same process for internal operations that do not cross a network boundary.

    Step 3: Configure the endpoint and token

    Connect the reporter to your Managed Service for OpenTelemetry endpoint and pass the authentication token.

    Replace the following placeholders with your actual values:

    PlaceholderDescription
    <endpoint>SkyWalking collector endpoint from the console
    <token>Authentication token from the console
    <service_name>A name that identifies your application
    <instance_name>A name that identifies this specific instance
    let endpoint = "<endpoint>";
let token = "<token>";
let service_name = "<service_name>";
let instance_name = "<instance_name>";

let reporter = GrpcReporter::connect(endpoint).await?;
let reporter = reporter.with_authentication(token);
let tracer = Tracer::new(service_name, instance_name, reporter.clone());

    For more information about how to obtain an endpoint and an authentication token, see Prerequisites.

    Step 4: Build and restart the application

    After you add instrumentation and configure the endpoint, rebuild and restart your application:

    cargo build --release
./target/release/<your-binary-name>

    FAQ

    Q: The build fails with a protobuf-related error.

    Build error

    A: protoc is not installed or not in your PATH. Fix this in one of two ways:

    • Install protoc using your OS package manager. See Prerequisites.

    • Add the vendored feature to the SkyWalking dependency so that protoc is compiled automatically during the build:

      skywalking = { version = "0.8.0", features = ["vendored"] }

    See also