ro11y-0.1.1 has been yanked.

ro11y

Lightweight Rust observability. Hand-rolled OTLP protobuf over HTTP, built on tracing.

What it does

Exports traces and logs as native OTLP protobuf over HTTP — no gRPC, no tonic, no prost
Dual output: OTLP HTTP to Vector + JSON stderr (local dev / CloudWatch fallback)
Deterministic trace IDs from CloudFront request IDs via BLAKE3
RED metrics (request duration, count, errors) as structured log events
W3C traceparent propagation for outbound requests
Tower middleware for Axum integration
Fire-and-forget with 3-retry exponential backoff — telemetry never blocks your service
7 direct dependencies, ~2000 lines of code

Usage

use ro11y::{init, TelemetryConfig};
use std::time::Duration;

let _guard = init(TelemetryConfig {
    service_name: "my-service",
    service_version: env!("CARGO_PKG_VERSION"),
    environment: "prod",
    otlp_endpoint: Some("http://vector:4318"),
    use_metrics_interval: Some(Duration::from_secs(30)),
});

// Tower middleware for Axum
let app = axum::Router::new()
    .route("/health", axum::routing::get(health))
    .layer(ro11y::request_layer())       // inbound: request spans + RED metrics
    .layer(ro11y::propagation_layer());  // outbound: W3C traceparent injection

Pipeline

Service (tracing) → ro11y (protobuf) → HTTP POST → Vector (OTLP source) → S3 Parquet

Why not OpenTelemetry SDK?

Version lock-step across opentelemetry-* crates
~120 transitive dependencies, 3+ minute compile times
Shutdown footgun (drop() doesn't flush)
gRPC bloat from tonic/prost

ro11y hand-rolls the protobuf wire format (~200 lines). The format has been stable since 2008.

License