avx-telemetry

Observability and tracing layer for Avila Experience Fabric

Production-grade observability for Rust applications. Integrates structured logging, distributed tracing, metrics collection, and scientific time series analysis from avila-telemetry.

Features

Structured logging: JSON and pretty-print formats
Distributed tracing: Span hierarchies with context propagation
Metrics integration: Performance counters and gauges
HTTP middleware: Automatic request tracing for axum/tower
Scientific analysis: Leverage avila-telemetry for time series anomaly detection
Multiple outputs: Console, file, OpenTelemetry, cloud services
Environment-aware: Development vs production configurations

Installation

[dependencies]
avx-telemetry = "0.1"

# For HTTP middleware
avx-telemetry = { version = "0.1", features = ["middleware"] }

Quick Start

Basic Setup

use avx_telemetry::init_tracing;

fn main() {
    // Initialize with default settings
    init_tracing();

    tracing::info!("Application started");
}

Structured Logging

use tracing::{info, warn, error};

#[tracing::instrument]
fn process_data(user_id: &str, count: usize) {
    info!(user_id, count, "Processing data");

    if count > 1000 {
        warn!(count, threshold = 1000, "High data volume");
    }
}

// Output (JSON):
// {"level":"INFO","fields":{"user_id":"123","count":500},"message":"Processing data"}

HTTP Middleware (Axum)

use axum::{Router, routing::get};
use avx_telemetry::middleware::TraceLayer;

#[tokio::main]
async fn main() {
    avx_telemetry::init_tracing();

    let app = Router::new()
        .route("/", get(handler))
        .layer(TraceLayer::new());

    axum::Server::bind(&"0.0.0.0:3000".parse().unwrap())
        .serve(app.into_make_service())
        .await
        .unwrap();
}

async fn handler() -> &'static str {
    "Hello, AVX!"
}

Each request automatically gets:

Request ID generation
Timing information
Status code tracking
Error logging

Time Series Analysis

Combine with avila-telemetry for scientific-grade analysis:

use avx_telemetry::TimeSeries;
use avila_telemetry::AnomalyDetector;

#[tracing::instrument]
fn analyze_metrics(values: Vec<f64>) {
    let ts = TimeSeries::new(values);
    let detector = AnomalyDetector::new(3.0, 1.5);

    if let Ok(anomalies) = detector.detect_zscore(&ts) {
        if !anomalies.is_empty() {
            tracing::warn!(
                count = anomalies.len(),
                "Anomalies detected in metrics"
            );
        }
    }
}

Configuration

Development Mode

use avx_telemetry::{TracingConfig, Format};

let config = TracingConfig {
    format: Format::Pretty,
    level: "debug",
    ..Default::default()
};

avx_telemetry::init_with_config(config);

Production Mode

let config = TracingConfig {
    format: Format::Json,
    level: "info",
    output_file: Some("logs/app.log"),
    ..Default::default()
};

Span Context

Create hierarchical traces:

use tracing::{info_span, info};

fn handle_request(user_id: &str) {
    let _span = info_span!("handle_request", user_id).entered();

    authenticate(user_id);
    fetch_data(user_id);
    process_response();
}

#[tracing::instrument]
fn authenticate(user_id: &str) {
    info!("Authenticating user");
}

Output shows parent-child relationship:

handle_request{user_id="123"}
  ├─ authenticate{user_id="123"}
  ├─ fetch_data{user_id="123"}
  └─ process_response

Metrics Collection

use avx_telemetry::metrics;

// Counters
metrics::counter!("requests_total", 1);

// Gauges
metrics::gauge!("active_connections", 42.0);

// Histograms
let start = std::time::Instant::now();
// ... do work ...
metrics::histogram!("request_duration_ms", start.elapsed().as_millis() as f64);

Integration with AVX Ecosystem

Works seamlessly with other AVX components:

use avx_gateway::Gateway;
use avx_telemetry::middleware::TraceLayer;

let gateway = Gateway::builder()
    .with_telemetry(TraceLayer::new())
    .build();

Features Flags

middleware: HTTP tracing middleware for axum/tower
metrics: Metrics collection and export

Part of AVX Ecosystem

avx-telemetry is part of the Avila Experience (AVX) platform:

avx-gateway: Uses telemetry for request tracing
avx-quantum-render: Performance diagnostics
avila-telemetry: Scientific time series analysis

Examples

See the examples/ directory:

cargo run --example basic_tracing
cargo run --example http_middleware --features middleware
cargo run --example time_series_integration

License

MIT OR Apache-2.0

See LICENSE-MIT and LICENSE-APACHE for details.

avx-telemetry 0.1.0