Appearance

@json-express/logger-pino ​

Enterprise-grade structured logger for JSONExpress using Pino.

The @json-express/logger-pino package implements ILogger using Pino, the fastest Node.js JSON logger. It supports both Twelve-Factor stdout streaming (for Docker/Kubernetes) and Spring Boot-style file-based logging (for traditional VM deployments).

Installation ​

bash
npm install @json-express/logger-pino pino-pretty

(Note: pino-pretty is optional but recommended for local development).

Configuration ​

bash
# .env

# Log level: trace | debug | info | warn | error | fatal
jex.log.level=info

# Output destination: "stdout", "stderr", or a file path
jex.log.path=stdout

# Human-readable formatting (auto-enabled in development)
jex.log.pretty=true

Logging Modes ​

1. Twelve-Factor Mode (Stdout) When jex.log.path=stdout, all logs are streamed to standard output as raw JSON. This is the recommended mode for Docker containers and Kubernetes Pods, where a centralized log aggregator (like Datadog, Fluentd, or CloudWatch) scrapes stdout.

json
{"level":30,"time":1714300800000,"traceId":"a1b2-c3d4","component":"API-REST","msg":"GET /albums 200 (12ms)"}

2. Enterprise File Mode When jex.log.path=./logs, Pino writes structured JSON logs to ./logs/app.log. The directory is automatically created if it doesn't exist (mkdir: true).

3. Pretty Mode (Development) When jex.log.pretty=true, the logger uses pino-pretty to render colorized, human-readable output with timestamps stripped of noise.

Core Features ​

1. Automatic Request Tracing ​

Every log entry automatically includes the traceId from RequestContext.getTraceId(). This means you can search your log aggregator for a single UUID and instantly see every database query, middleware decision, and hook execution that occurred during that specific HTTP request.

2. Child Loggers ​

The .child() method creates scoped logger instances. Every JSONExpress plugin receives a child logger with the component name pre-injected:

typescript
const dbLogger = logger.child({ component: 'Adapter-Postgres' });
dbLogger.info('Connected to database'); 
// → { component: "Adapter-Postgres", traceId: "...", msg: "Connected to database" }

3. Zero-Overhead in Production ​

Pino is designed to add virtually zero overhead to your application. Unlike Winston or Bunyan, Pino uses worker threads for serialization and avoids synchronous I/O entirely.