Monitoring & Observability

Level 1: Quick Reference

Three Pillars of Observability

Metrics - Numerical measurements over time

• Counter (only increases): request_total, errors_total
• Gauge (can go up/down): cpu_usage, memory_bytes
• Histogram (distribution): request_duration_seconds
• Summary (quantiles): response_time_summary

Logs - Timestamped event records

• Structured (JSON): {"level":"error","msg":"connection failed","user_id":123}
• Unstructured (text): 2025-01-15 ERROR: Connection timeout
• Log levels: DEBUG, INFO, WARN, ERROR, FATAL