This section describes the architecture of the Observability subsystem of Deckhouse Kubernetes Platform (DKP).

The Observability subsystem includes the following modules:

  • prometheus: Deploys a monitoring stack with predefined settings for DKP and applications, simplifying the initial configuration.
  • operator-prometheus: Installs Prometheus Operator, which automates the deployment and management of Prometheus instances.
  • prometheus-metrics-adapter: Allows HPA and VPA autoscalers to use monitoring metrics when making scaling decisions.
  • log-shipper: Simplifies log collection setup in Kubernetes clusters.
  • loki: Deploys a short-term log storage system in the cluster based on Grafana Loki.
  • observability: Extends the functionality of the prometheus module and the Deckhouse web UI, providing additional capabilities for flexible management of metrics, dashboards, and alerts, as well as access control mechanisms for them.
  • extended-monitoring: Enhances cluster monitoring by deploying additional Prometheus exporters that help detect potential issues before they affect service operation.
  • monitoring-custom: Simplifies monitoring configuration for user applications by requiring only a specific label to be set for the target application.
  • monitoring-deckhouse: Provides monitoring of DKP components and services.
  • monitoring-kubernetes: Provides transparent and timely monitoring of all cluster nodes and key infrastructure components.
  • monitoring-kubernetes-control-plane: Organizes secure metrics collection and provides a basic set of monitoring rules for cluster control plane components.
  • upmeter: Checks platform availability and cluster component health in real time and displays the results on dedicated dashboards.