📊 Structured Logging & Sentry Integration

Production observability begins with structured logging. Plain text println! statements are difficult to search in log aggregators like Datadog, Grafana Loki, or AWS CloudWatch.

Ferrox provides ferrox-logger—a crate that initializes structured JSON logging via tracing-subscriber and configures automatic error reporting via Sentry.

---

1. Initializing the Logger

Call setup_logger at the beginning of main.rs:

use ferrox_logger::{setup_logger, LoggerConfig};
use tracing::{info, warn, error};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let mut config = LoggerConfig::default();
    config.service_name = "my-service".to_string();
    config.environment = "production".to_string();

    // Initializes tracing subscriber and optional Sentry guard
    let _sentry_guard = setup_logger(config)?;

    info!("🚀 Application initialized with structured logging!");
    Ok(())
}

---

2. Emitting Structured Log Spans

Use the standard tracing macros (info!, warn!, error!, debug!) to attach key-value metadata to logs:

use tracing::info;

pub async fn process_order(order_id: u64, user_id: u64, amount: f64) {
    info!(
        order_id = order_id,
        user_id = user_id,
        amount = amount,
        "Order processing started"
    );
}

Output JSON log line:

{
  "timestamp": "2026-09-05T10:00:00Z",
  "level": "INFO",
  "fields": {
    "order_id": 42,
    "user_id": 101,
    "amount": 99.95,
    "message": "Order processing started"
  },
  "target": "my_service::orders"
}

---

3. Sentry Integration

When LoggerConfig.sentry_dsn is configured, all error! events automatically create detailed Sentry exception reports complete with stack traces, environment tags, and request contexts.