Prometheus ecosystem components¶
The Prometheus ecosystem comprises several distinct components that work together to collect, process, and visualize Metrics. When deploying Prometheus in a Kubernetes environment using the official community Helm chart, these components are typically installed as a unified stack.^[prometheus.md]
Core Components¶
The installation creates several key resources within the cluster:
- Prometheus Server: The central monitoring service that scrapes and stores Metrics data.^[prometheus.md]
- Alertmanager: A dedicated component for handling alerts, grouping them, and routing notifications to defined receivers.^[prometheus.md]
- Node Exporter: A deployment (often running as a DaemonSet) that exposes hardware and OS-level Metrics from cluster nodes.^[prometheus.md]
- Kube State Metrics: A service that generates Metrics about Kubernetes objects, such as pods and nodes, translating the cluster state into Prometheus-compatible Metrics.^[prometheus.md]
- Pushgateway: A component for collecting Metrics from short-lived jobs, allowing these jobs to push their Metrics to an intermediary that Prometheus then scrapes.^[prometheus.md]
Visualization¶
While not part of the Prometheus software itself, Grafana is the standard component used for visualization in the ecosystem.^[prometheus.md]
Grafana connects to the Prometheus data source to query and display the collected Metrics.^[prometheus.md] In Kubernetes deployments, Grafana communicates with the Prometheus server using the internal ClusterIP service (e.g., http://prometheus-server:80).^[prometheus.md]
Related Concepts¶
- Prometheus
- Kubernetes
- [[Monitoring]]
Sources¶
^[prometheus.md]