Overview
Monitoring is the practice of collecting, aggregating, and analyzing metrics, logs, and traces from systems to understand their health and performance. It provides the foundation for alerting, capacity planning, and incident response. Effective monitoring distinguishes between leading indicators (queue depth, error rate) and lagging indicators (customer complaints, revenue impact). Modern monitoring stacks typically combine time-series databases, log aggregators, and dashboarding tools to provide unified visibility across infrastructure and application layers.
Key Concepts
- Metrics — Numerical measurements collected at regular intervals (CPU usage, request latency, error counts). Stored in time-series databases like Prometheus or InfluxDB.
- Alerting — Rules that trigger notifications when metrics cross defined thresholds. Good alerts are actionable, not noisy.
- Dashboards — Visual representations of system health. Effective dashboards answer “is the system healthy?” at a glance.
- The Four Golden Signals — Latency, traffic, errors, and saturation (from Google SRE). These four metrics capture the essential health of any service.
- Log aggregation — Centralizing logs from distributed systems for searchability and correlation.
- Health checks — Periodic probes that verify service availability and readiness.
Trade-offs
- Too many alerts cause alert fatigue; too few miss real incidents
- High-cardinality metrics provide detail but increase storage and query costs
- Sampling reduces overhead but may miss rare events
Related
- Observability extends monitoring with deeper diagnostic capability
- Site Reliability Engineering relies on monitoring to define and measure SLOs
- CD pipelines benefit from deployment monitoring