Skip to content

Installation

This page provides comprehensive instructions for installing the Qubership Monitoring Operator into Kubernetes or OpenShift environments. The documentation covers prerequisites, supported versions, installation procedures, and post-installation verification.

Supported Versions

The Qubership Monitoring Operator follows the N±2 Kubernetes version support policy. Currently, the recommended Kubernetes version is 1.26.x.

Kubernetes version Support Status
1.24.x Tested
1.25.x Tested
1.26.x (recommended) Tested
1.27.x Forward compatible
1.28.x Forward compatible

For OpenShift compatibility, refer to the equivalent Kubernetes version. For example, OpenShift 4.12 is based on Kubernetes 1.25.0.

Cloud Platform Support:

Cloud Platform Support
AWS Elastic Kubernetes Service (EKS) ✓
Azure Kubernetes Service (AKS) ✓
Google Kubernetes Engine (GKE) ✓
On-premise Kubernetes >= 1.25 ✓
On-premise OpenShift >= 4.10 ✓

Prerequisites

Before You Begin

Ensure you have the following prerequisites in place before installation:

  • Kubernetes 1.19+ or OpenShift 3.11+ cluster
  • kubectl 1.19+ or oc 3.11+ CLI tools
  • Helm 3.0+
  • Pre-created namespace for installation
  • Appropriate permissions for deployment
  • Sufficient hardware resources

Important considerations:

  • Apply CRDs manually if you disable automated CRD upgrade
  • For OpenShift deployments, ensure the namespace has an empty node-selector

For detailed prerequisites, see Prerequisites.

Permissions

The monitoring operator requires cluster-level permissions to create and manage the following components:

  • prometheus
  • prometheus-operator
  • grafana-operator
  • kube-state-metrics
  • node-exporter
  • cert-exporter
  • Various exporters and monitors

A ClusterRole must be granted to the monitoring-operator service account with all required permissions. Alternatively, the operator can be deployed with restricted privileges, but additional manual setup is required.

Hardware Requirements

The monitoring stack requires resources depending on the components installed and the scale of your environment.

Minimal recommended hardware for an average system:

Resource Requirement
CPU 3000m
RAM 3000Mi
Storage 10GB

For detailed hardware sizing information, refer to the Prerequisites guide.

Default Deployment

By default, the following components are installed:

Component Description
alertmanager Alert management and routing
grafana Visualization and dashboards
grafana-operator Grafana management
kube-state-metrics Kubernetes metrics collection
monitoring-operator Core operator (always installed)
node-exporter Node-level metrics
victoriametrics Time series database
vm-operator VictoriaMetrics management

Default configuration includes:

  • Ephemeral storage for VictoriaMetrics and Grafana
  • 24-hour metric retention
  • Multiple pre-configured dashboards and alerts
  • No ingress configured by default
  • No integrations enabled

For detailed information about components, see Components.

Installation Steps

Using Helm

  1. Add the Helm repository: 

    helm repo add qubership-monitoring https://your-repo.com/
helm repo update

  2. Create values file (values.yaml) with your customizations or use the default values.

  3. Install the chart: 

    helm install monitoring qubership-monitoring/monitoring-operator \
  --namespace monitoring \
  --create-namespace \
  --values values.yaml

  4. Upgrading the chart: 

    helm upgrade monitoring qubership-monitoring/monitoring-operator \
  --namespace monitoring \
  --values values.yaml

  5. Uninstalling the chart: 

    helm uninstall monitoring --namespace monitoring

CRD Management

Helm does not update or remove CRDs. For upgrades involving CRD changes, manual updates are required.

Post-Installation Verification

After installing the monitoring operator, verify that all components are running correctly:

  1. Check pod status: 

    kubectl get pods -n monitoring

  2. Verify the operator deployment: 

    kubectl get deployment monitoring-operator -n monitoring

  3. Check Custom Resources: 

    kubectl get crd | grep monitoring

  4. Access Grafana dashboard:

If ingress is configured, access via URL. Otherwise, use port-forwarding: 

kubectl port-forward svc/grafana 3000:3000 -n monitoring

Default login for Grafana is typically admin/admin unless configured otherwise.

For complete verification procedures, see Post-Deploy Checks.

What's Next

After successful installation, you can: