What are the three pillars of observability, and what is each good for?
What they are really testing: Whether you know metrics, logs, and traces and, crucially, what each is best at and how they complement each other rather than overlap.
A real interview question
What are the three pillars of observability, and what is each good for?
What most people say
drag me
“Metrics, logs, and traces, the three ways to monitor a system.”
It names them but not what each is best at or how they differ. Knowing that metrics are for trends/alerting, logs for detail, and traces for distributed latency is what makes the answer useful.
The follow-ups they ask next
Why not just use logs for everything?
Logs are expensive and noisy at scale and poor for aggregation/trends and for following a request across services. Metrics are far cheaper for trends/alerting; traces are purpose-built for distributed latency. Each pillar fits a different job.
When is a trace more useful than a metric?
When you need to know where in a multi-service request the time or error is, metrics show the overall latency rose, a trace shows which downstream span caused it.
What the interviewer is listening for
- Names all three with their strengths
- Knows logs are expensive, metrics cheap, traces for distributed latency
- Knows they compose
What sinks the answer
- Only lists them
- Cannot say what each is best at
- Thinks they are redundant
If you genuinely do not know
Say this instead of freezing. Reasoning out loud from what you do know beats silence every single time, and a good interviewer is listening for exactly that.
“Three pillars: [metrics (cheap numeric time series, for trends/dashboards/alerting)], [logs (detailed timestamped events, for the why, but expensive at scale)], [traces (a request across services in spans, for where time goes)]. They compose: [metrics = what/when, traces = where, logs = why], and you move between them in an investigation.”
Keep going with observability
Foundation
What is the difference between monitoring and observability?
Foundation
What should you measure first for a service? Explain the golden signals (or RED/USE).
Junior
What are good logging practices for a distributed system?
Junior
What are the main metric types (counter, gauge, histogram), and when do you use each?
Junior
What is distributed tracing, and how does it work?
Junior
How does Prometheus collect metrics, and what is the pull model?
Knowing the answer is not the same as recalling it under pressure
Sign in to send the questions you fumble to spaced recall, so they come back right before you would forget them, and learn the concepts behind them with hands-on labs.
Start free