Problem
Monitor a Kubernetes cluster
I need to collect metrics, build dashboards and alert on the health of a Kubernetes cluster and its workloads.
Cluster monitoring typically combines a metrics backend, a visualization layer and alerting. The projects below are commonly combined rather than used in isolation.
Candidates we have reviewed
Ordered by editorial rank for this problem. Rationale and caveats are editorial analysis; see each project page for the underlying evidence.
Prometheus
Open-source monitoring system and time-series database with a powerful query language (PromQL).
Why it fits: Scrapes Kubernetes metrics and evaluates alerting rules.
Grafana
Analytics and interactive visualization platform for metrics, logs and traces.
Why it fits: Dashboards and alerting on top of Prometheus data.
Caveat: AGPLv3 license since 2021.
OpenTelemetry Collector
Vendor-agnostic agent to receive, process and export telemetry (traces, metrics, logs).
Why it fits: Collects and routes metrics, traces and logs vendor-neutrally.
This is an editorial evaluation with citations, not a sponsored list. See ourmethodology.