Lesson 41 • Advanced

Logging with SLF4J & Logback

System.out.println() is for learning. Professional logging is for production. When your app crashes at 2 AM, your logs are the only witness. SLF4J + Logback gives you structured, searchable, level-filtered logs that actually help you debug — fast.

Before You Start

You should know Exception Handling (logging exceptions properly) and Maven & Gradle (adding logging dependencies). Understanding Multithreading helps for MDC context.

What You'll Learn

✅ SLF4J as a logging facade
✅ Log levels: TRACE, DEBUG, INFO, WARN, ERROR
✅ Logback configuration and appenders
✅ MDC (Mapped Diagnostic Context) for request tracing
✅ Structured logging (JSON format)
✅ Log rotation and production best practices

1️⃣ Log Levels — When to Use Each

Key insight: In production, you typically run at INFO level. TRACE and DEBUG messages are silently discarded (zero cost with parameterized logging). When investigating issues, temporarily lower to DEBUG for a specific package without restarting.

Level	When to Use	Production
TRACE	Method entry/exit, loop iterations	Off
DEBUG	Variable values, algorithm steps	Off
INFO	App started, user logged in, order placed	On ✅
WARN	Retry attempt, deprecated API, slow query	On ✅
ERROR	Unhandled exceptions, failed payments	On ✅ + alerts

Try It: Logging System Simulation

Try it Yourself »

JavaScript

// 💡 Try modifying this code and see what happens!
// Professional logging system simulation

console.log("=== Logging System ===\n");

// Simulated SLF4J Logger
const LEVELS = { TRACE: 0, DEBUG: 1, INFO: 2, WARN: 3, ERROR: 4 };
let currentLevel = LEVELS.INFO; // Production default

class Logger {
  constructor(name) { this.name = name; }
  
  log(level, msg, ...args) {
    if (LEVELS[level] < currentLevel) return; // Filtered out!
    let formatted = msg.replace(/\{\}/g, () => args.shift() ?? 
...

2️⃣ MDC & Structured Logging

Analogy: MDC is like a name badge that every log message from a request automatically wears. When 1000 requests are processing simultaneously, MDC lets you filter logs by requestId to see exactly one request's journey through your system.

MDC.put("requestId", uuid) — attach to thread context

%X{requestId} — include in log pattern

MDC.clear() — clean up in finally block

Try It: MDC & Structured JSON Logging

Try it Yourself »

JavaScript

// 💡 Try modifying this code and see what happens!
// MDC request tracing and structured JSON logging

console.log("=== MDC & Structured Logging ===\n");

// Simulated MDC (thread-local context)
let MDC = {};

class StructuredLogger {
  constructor(name) { this.name = name; }
  
  log(level, msg) {
    // Plain text format
    let plain = new Date().toISOString().substr(11, 12) + " " + level.padEnd(5) + " " + 
      (MDC.requestId ? "[" + MDC.requestId + "] " : "") +
      (MDC.userId ? "user="
...

Common Mistakes

⚠️ String concatenation: log.debug("User " + name) wastes CPU. Use log.debug("User {}", name)
⚠️ Logging sensitive data: Never log passwords, credit cards, tokens, or PII
⚠️ Catching and swallowing: catch(Exception e) {} is invisible failure. Always log the exception
⚠️ Using System.out.println: Not filtered, not structured, not searchable. Use SLF4J

Pro Tips

💡 MDC — attach requestId/userId to every log. Essential for microservice tracing
💡 Structured JSON logging — log as JSON so Elasticsearch/CloudWatch can parse and search
💡 Lombok's @Slf4j — one annotation replaces the logger boilerplate
💡 Log rotation — configure max file size and history to prevent disk fill

Try It: Log Rotation & Alerting

Try it Yourself »

JavaScript

// 💡 Try modifying this code and see what happens!
// Log rotation and alerting simulation

console.log("=== Log Rotation & Alerting ===\n");

// 1. Log rotation configuration
console.log("1. LOG ROTATION CONFIG:");
console.log(`  <appender name="FILE" class="RollingFileAppender">
    <file>logs/app.log</file>
    <rollingPolicy class="SizeAndTimeBasedRollingPolicy">
      <fileNamePattern>logs/app.%d{yyyy-MM-dd}.%i.log.gz</fileNamePattern>
      <maxFileSize>100MB</maxFileSize>
      <maxHisto
...

📋 Quick Reference

Concept	API	Purpose
Get logger	LoggerFactory.getLogger(X.class)	SLF4J factory
Parameterized	log.info("msg {}", val)	Lazy string eval
MDC	MDC.put("key", "value")	Thread-local context
Markers	MarkerFactory.getMarker("AUDIT")	Categorize entries
Lombok	@Slf4j	Auto-generate logger

🎉 Lesson Complete!

You've mastered professional logging — your logs will now be structured, searchable, and actionable! Next: Secure Coding — preventing vulnerabilities in Java applications.

Logging with SLF4J & Logback

Before You Start

What You'll Learn

1️⃣ Log Levels — When to Use Each

Try It: Logging System Simulation

2️⃣ MDC & Structured Logging

Try It: MDC & Structured JSON Logging

Common Mistakes

Pro Tips

Try It: Log Rotation & Alerting

📋 Quick Reference

🎉 Lesson Complete!

Cookie & Privacy Settings