Basic Components¶

This document describes the monitoring components and their deployment scenarios.

Component Overview¶

The Qubership Monitoring Operator includes many components that serve various monitoring functions. Not all projects need all functions, so you can specify which components to install.

Always Installed¶

monitoring-operator - Core operator that manages all other components

Default Configuration¶

By default, the following components are installed:

monitoring-operator - Core operator
alertmanager - Alert management and routing
grafana - Visualization and dashboards
grafana-operator - Grafana management
kube-state-metrics - Kubernetes state metrics collection
node-exporter - Node-level system metrics
victoriametrics - Time series database and storage
vm-operator - VictoriaMetrics management

Available Components¶

Core Monitoring Stack¶

VictoriaMetrics Stack¶

vmoperator - VictoriaMetrics operator
vmagent - Metrics collection agent
vmalert - Alerting component
vmalertmanager - Alert management
vmauth - Authentication proxy
vmsingle - Single-node VictoriaMetrics
vmuser - User management

Grafana Stack¶

grafana-operator - Grafana instance management
grafana - Visualization and dashboard platform
grafana-dashboards - Pre-built monitoring dashboards

AlertManager¶

alertmanager - Alert routing, grouping, and notification management

Prometheus Stack¶

prometheus-operator - Prometheus instance management
prometheus - Time series database and monitoring engine
prometheus-rules - Alerting and recording rules
kubernetes-monitors - Kubernetes-specific monitoring targets

Metrics Exporters¶

Core Exporters¶

node-exporter - Hardware and OS metrics
kube-state-metrics - Kubernetes object state metrics

Specialized Exporters¶

blackbox-exporter - External endpoint probing
cert-exporter - TLS certificate monitoring
version-exporter - Version tracking
network-latency-exporter - Network latency measurements

Cloud Exporters¶

cloudwatch-exporter - AWS CloudWatch metrics
promitor-agent-scraper - Azure Monitor metrics
stackdriver-exporter - Google Cloud metrics

Additional Components¶

pushgateway - Push-based metrics collection
promxy - Prometheus proxy and aggregator
prometheus-adapter - Kubernetes custom metrics API
prometheus-adapter-operator - Prometheus adapter management
graphite-remote-adapter - Graphite protocol support

Deployment Scenarios¶

Scenario 1: Metrics Collection Only¶

Minimal setup for basic metrics collection:

components:
  - monitoring-operator
  - node-exporter
  - kube-state-metrics
  - victoriametrics

Scenario 2: Collection + Visualization¶

Add dashboards and visualization:

components:
  - monitoring-operator
  - node-exporter
  - kube-state-metrics
  - victoriametrics
  - grafana
  - grafana-operator

Scenario 3: Full Monitoring Stack¶

Complete monitoring with alerting:

components:
  - monitoring-operator
  - alertmanager
  - grafana
  - grafana-operator
  - kube-state-metrics
  - node-exporter
  - victoriametrics
  - vm-operator

Scenario 4: External Storage Integration¶

Send metrics to external systems:

components:
  - monitoring-operator
  - node-exporter
  - kube-state-metrics
  - vmagent  # Collection agent
  - graphite-remote-adapter  # External storage

Default Deployment Settings¶

Resource Allocation¶

Default resource requests and limits:

Component	Requests	Limits
`alertmanager`	CPU: 100m, RAM: 100Mi	CPU: 200m, RAM: 200Mi
`grafana`	CPU: 300m, RAM: 400Mi	CPU: 500m, RAM: 800Mi
`grafana-operator`	CPU: 50m, RAM: 50Mi	CPU: 100m, RAM: 100Mi
`kube-state-metrics`	CPU: 50m, RAM: 50Mi	CPU: 100m, RAM: 256Mi
`monitoring-operator`	CPU: 50m, RAM: 50Mi	CPU: 100m, RAM: 150Mi
`node-exporter`	CPU: 50m, RAM: 50Mi	CPU: 100m, RAM: 100Mi
`victoriametrics`	CPU: 1000m, RAM: 3Gi	CPU: 1500m, RAM: 5Gi
`vm-operator`	CPU: 100m, RAM: 100Mi	CPU: 200m, RAM: 200Mi

Default Behavior¶

Storage: VictoriaMetrics and Grafana use ephemeral storage (emptyDir)
Retention: VictoriaMetrics stores metrics for 24 hours by default
Ingress: No ingress configured by default (Kubernetes)
Integrations: No external integrations enabled
Exporters: Basic exporters only (no blackbox/cert exporters)

Metrics Collection Sources¶

By default, metrics are collected from:

Kubelet and cAdvisor
Kubernetes API Server
Nginx Ingress (if available)
Etcd
Kube-state-metrics
Node Exporter
Self-monitoring: VictoriaMetrics, AlertManager, Grafana

Pre-built Resources¶

Dashboards: See Default Dashboards
Alerts: See Default Alerts

Component Configuration¶

Enabling/Disabling Components¶

You can selectively enable or disable:

Dashboards¶

Configure via grafanaDashboards.list parameter - specify which dashboards to install.

Prometheus Rules¶

Configure via prometheusRules.ruleGroups parameter - specify alert rule groups to install.

Kubernetes Monitors¶

Configure via kubernetesMonitors parameter - specify monitoring targets and override scraping parameters.

Automatic Rules¶

Some dashboards and monitors are automatically enabled/disabled based on component availability:

Dashboard Rules¶

kubernetes-nodes-resources - disabled if node-exporter not installed
home-dashboard - disabled if grafana.grafanaHomeDashboard is false
core-dns-dashboard - disabled if CoreDNS monitor not available or OpenShift ≤ 3.11
Ingress dashboards - disabled on OpenShift or if nginx ingress monitor unavailable
VictoriaMetrics dashboards - enabled only with VM components
Prometheus dashboards - enabled only with Prometheus components

Example Configurations¶

Basic Setup¶

# Minimal installation - use all defaults

Custom Ingress Setup¶

grafana:
  ingress:
    install: true
    host: grafana.example.com

victoriametrics:
  vmsingle:
    install: true
    ingress:
      install: true
      host: victoriametrics.example.com

  vmAlertManager:
    install: true  
    ingress:
      install: true
      host: alertmanager.example.com

Cloud-Specific Setup¶

# AWS with CloudWatch
cloudwatch-exporter:
  install: true

# Azure with Promitor
promitor-agent-scraper:
  install: true

# GCP with Stackdriver
stackdriver-exporter:
  install: true

Next Steps¶

Review Storage configuration options
Plan your Deployment strategy
Configure individual Components as needed