A Simple Analogy

Logging is like keeping a detailed diary. Every important event gets recorded for later analysis.

---

Why Logging?

• Debugging: Track application flow
• Monitoring: Detect issues in production
• Auditing: Record user actions
• Performance: Identify bottlenecks
• Support: Help diagnose customer issues

---

Setup Serilog

var logger = new LoggerConfiguration()
    .MinimumLevel.Debug()
    .WriteTo.Console()
    .WriteTo.File("logs/app-.txt", 
        rollingInterval: RollingInterval.Day)
    .Enrich.FromLogContext()
    .Enrich.WithProperty("Environment", env)
    .CreateLogger();

Log.Logger = logger;
builder.Host.UseSerilog(logger);

---

Structured Logging

// Bad
logger.LogInformation("User logged in");

// Good
logger.LogInformation("User {UserId} logged in at {LoginTime} from {IpAddress}",
    userId, DateTime.UtcNow, ipAddress);

// Output: User 123 logged in at 2025-02-20 10:30:00 from 192.168.1.1

---

Log Levels

logger.LogDebug("Debug: Detailed information");
logger.LogInformation("Info: General information");
logger.LogWarning("Warning: Potential problem");
logger.LogError("Error: Something failed");
logger.LogCritical("Critical: System failure");

---

Exception Logging

try
{
    await ProcessOrderAsync(order);
}
catch (Exception ex)
{
    logger.LogError(ex, "Failed to process order {OrderId}", order.Id);
    throw;
}

---

Best Practices

1. Structured: Include context in logs
2. Levels: Use appropriate severity
3. Performance: Don't log sensitive data
4. Context: Add correlation IDs
5. Aggregation: Use log aggregation tools

---

Related Concepts

• Log aggregation (ELK, Splunk)
• Application Insights
• CloudWatch
• Distributed tracing

---

Summary

Implement structured logging with Serilog. Include context information and use appropriate log levels for effective monitoring.