Monitoring

Definitions

1. Monitoring is the practice of continuously collecting, aggregating, and analyzing metrics, logs, and health signals from applications and infrastructure to detect problems, track performance, and trigger alerts before issues affect users. Typical metrics include CPU usage, memory consumption, request latency, error rates, and throughput. Tools like Prometheus, Grafana, Datadog, and CloudWatch provide dashboards and alerting pipelines. Monitoring differs from observability: monitoring tells you that something is wrong, while observability helps you understand why. Effective monitoring includes setting service-level indicators (SLIs), defining thresholds and alert rules, and avoiding alert fatigue by tuning sensitivity to focus on actionable signals.

   In plain English: Keeping a constant eye on your servers and apps so you know when something breaks before your customers do.

   Example: "If you don't have monitoring, you find out about outages from your users on Twitter."

Related Terms

• distributed tracing
• runbook
• toil
• configuration drift
• Incident Management
• Logging
• Seq
• Data Lineage
• Backfill
• OpenTelemetry
• Runbook
• Connection Leak
• Jaeger
• Metrics Collection
• Tracing
• BPF Security