Observability Stack: Prometheus, Grafana, and Loki in Kubernetes
Metrics, logs, and traces — the three pillars of observability. We present our proven stack for Kubernetes environments.
The Three Pillars of Observability
Monitoring tells you that something is broken. Observability tells you why. For that, you need three data sources: metrics, logs, and traces.
Prometheus: Metrics
Prometheus is the de facto standard for Kubernetes metrics. It scrapes metrics from your applications and infrastructure at regular intervals.
- ServiceMonitor: Declarative configuration of which services are monitored — no manual scrape configs.
- AlertManager: Routing of alerts to Slack, PagerDuty, or email based on severity and team.
- PromQL: Powerful query language for metrics — from simple gauges to complex rate calculations.
Grafana: Visualization
Grafana unifies all data sources in a single dashboard. From Kubernetes cluster health and application metrics to business KPIs — everything in one place.
Loki: Logs
Loki is log aggregation for Kubernetes. Unlike ELK, Loki does not index log content, only labels — making it faster and more cost-effective.
- Promtail: An agent on every node that automatically sends container logs to Loki.
- LogQL: A query language modeled after PromQL — a consistent experience for teams familiar with Prometheus.
- Grafana Integration: View logs alongside metrics in the same dashboard — correlate spikes with log entries.
Our Recommendation
The PLG stack (Prometheus, Loki, Grafana) is the standard for every Kubernetes project at devRocks. It is open source, battle-tested, and scales from small clusters to enterprise environments.
Questions About This Topic?
We are happy to advise you on the technologies and solutions described in this article.
Get in Touch
