Many developers think observability means "add more logs."
Logging is important.
But observability is broader.
The Three Common Pillars
Observability is commonly discussed through:
Logs Metrics Traces Each answers different questions.
Logs
Logs tell you about events.
Example:
2026-08-14 14:20:03 Order 12345 payment failed Useful, but often isolated.
Metrics
Metrics measure numerical behavior.
For example:
HTTP requests/sec CPU usage Memory usage Error rate Request latency You might discover:
95th percentile latency = 2.4 seconds That tells you something is happening at scale.
Traces
Distributed tracing follows a request across multiple components.
For example:
API Gateway ↓ User Service ↓ Order Service ↓ Payment Service ↓ Database If the request takes five seconds, tracing can help identify where that time went.
Why Logs Alone Aren't Enough
Imagine:
Payment failed A log tells you the payment failed.
A metric tells you that payment failures increased from:
0.5% → 12% A trace can help show that requests are timing out inside a specific dependency.
Together, these signals provide much better visibility.
Final Thoughts
Logging answers:
What happened?
Metrics help answer:
How often is it happening?
Tracing helps answer:
Where did the request spend its time?
Understanding all three is an important step toward production-grade engineering.
