Monitoring standalone Kubernetes clusters Innovation Release

Use the EDB Postgres AI agent for centralized monitoring, metrics collection, and resource management for your EDB Postgres® AI for CloudNativePG™ or CloudNativePG™ clusters and integrate them on your Hybrid Manager (HM) console. By integrating these clusters with your HM console, you gain a unified view of your cloud-native environment. This integration is deployed using a Helm chart.

You can configure the EDB Postgres AI agent to run anywhere. To activate specific capabilities, assign one or more roles to it: Webhooks, Providers, Dispatchers, or Runners. You can configure a single agent to perform any combination of these functions simultaneously.

Architecture

The architecture to implement monitoring Kubernetes deployments consists of the following components.

EDB Postgres AI agent

The EDB Postgres AI agent serves as a centralized tool for resource discovery and monitoring, while also streamlining the collection and forwarding of your metrics.

OpenTelemetry collector (optional)

The OpenTelemetry (OTEL) collector acts as a centralized data pipeline that gathers logs and exports telemetry data from your environment.

Key functions:

• Container metrics (cAdvisor): The cAdvisor service collects and processes resource usage data (CPU, memory) from active containers.
• Kubelet metrics: Monitors the health and performance of Kubernetes nodes and pods.
• Application logs: Aggregates and forwards logs from your database applications.
• Custom metrics: Allows for the collection of user-defined metrics tailored to specific application needs.

Kube-state-metrics (optional)

Kube-state-metrics generates metrics based on the state of Kubernetes objects. Enable it as a subchart with additionalCharts.kube-state-metrics.install: true. See the Configuration reference.