Skip to content

Latest commit

 

History

History
590 lines (474 loc) · 13.1 KB

File metadata and controls

590 lines (474 loc) · 13.1 KB
title Cluster Configuration Examples
category cluster-management
sources
title url date_accessed
Installing a single-node OpenShift cluster
2026-03-31
tags
examples
cluster-configurations
use-cases
reference-architectures
semantic_keywords
cluster configuration examples
real-world configurations
sno examples
ha cluster examples
deployment scenarios
use_cases
cluster-creator
configuration-reference
architecture-planning
related_docs
platforms.md
providers.md
networking.md
host-requirements.md
last_updated 2026-03-31

Cluster Configuration Examples

Real-world examples of valid OpenShift cluster configurations for different use cases.


Example 1: SNO for Edge Deployment

Scenario: Single-server edge location (retail store, factory floor)

Configuration:

cluster_name: edge-store-01
cluster_type: SNO
platform: none
openshift_version: 4.18.2
base_domain: retail.example.com
cpu_architecture: x86_64

Host Requirements:

  • 1 physical server
  • 8 CPU cores
  • 32 GB RAM
  • 120 GB SSD

Network Configuration:

networking_mode: DHCP
vips: Not required (SNO)

DNS Records Required:

api.edge-store-01.retail.example.com.  A  <dhcp-assigned-ip>
*.apps.edge-store-01.retail.example.com.  A  <dhcp-assigned-ip>

Use Case:

  • Point-of-sale applications
  • Local inventory management
  • Edge analytics
  • Disconnected or intermittent connectivity

Installation Time: ~45 minutes


Example 2: HA Cluster on Bare Metal (DHCP)

Scenario: On-premises data center with DHCP available

Configuration:

cluster_name: production-dc1
cluster_type: HA
platform: baremetal
openshift_version: 4.18.2
base_domain: datacenter.example.com
cpu_architecture: x86_64

Host Requirements:

  • 3 control plane nodes: 4 CPU, 16 GB RAM, 120 GB disk each
  • 3 worker nodes: 8 CPU, 32 GB RAM, 500 GB disk each

Network Configuration:

networking_mode: DHCP
machine_network: 192.168.100.0/24
api_vip: 192.168.100.10
ingress_vip: 192.168.100.11

DNS Records Required:

api.production-dc1.datacenter.example.com.  A  192.168.100.10
*.apps.production-dc1.datacenter.example.com.  A  192.168.100.11

Use Case:

  • Production applications
  • Corporate workloads
  • Traditional data center

Installation Time: ~60-90 minutes


Example 3: HA Cluster on Bare Metal (Static IPs)

Scenario: On-premises production cluster without DHCP

Configuration:

cluster_name: prod-ocp
cluster_type: HA
platform: baremetal
openshift_version: 4.18.2
base_domain: prod.example.com
cpu_architecture: x86_64

Host Requirements:

  • 3 control plane nodes
  • 2 worker nodes

Network Configuration (Static):

networking_mode: Static
machine_network: 10.50.0.0/24
api_vip: 10.50.0.100
ingress_vip: 10.50.0.101

static_network_configs:
  - host: master-0
    interface: ens3
    mac_address: "52:54:00:aa:bb:01"
    ipv4_address: 10.50.0.10
    prefix: 24
    gateway: 10.50.0.1
    dns_servers: ["10.50.0.5", "8.8.8.8"]

  - host: master-1
    interface: ens3
    mac_address: "52:54:00:aa:bb:02"
    ipv4_address: 10.50.0.11
    prefix: 24
    gateway: 10.50.0.1
    dns_servers: ["10.50.0.5", "8.8.8.8"]

  - host: master-2
    interface: ens3
    mac_address: "52:54:00:aa:bb:03"
    ipv4_address: 10.50.0.12
    prefix: 24
    gateway: 10.50.0.1
    dns_servers: ["10.50.0.5", "8.8.8.8"]

  - host: worker-0
    interface: ens3
    mac_address: "52:54:00:aa:bb:04"
    ipv4_address: 10.50.0.20
    prefix: 24
    gateway: 10.50.0.1
    dns_servers: ["10.50.0.5", "8.8.8.8"]

  - host: worker-1
    interface: ens3
    mac_address: "52:54:00:aa:bb:05"
    ipv4_address: 10.50.0.21
    prefix: 24
    gateway: 10.50.0.1
    dns_servers: ["10.50.0.5", "8.8.8.8"]

Use Case:

  • Regulated environments requiring static IPs
  • Networks without DHCP infrastructure
  • Predictable IP allocation requirements

Installation Time: ~90 minutes (additional time for static network config)


Example 4: HA Cluster on VMware vSphere

Scenario: Enterprise VMware environment

Configuration:

cluster_name: vsphere-prod
cluster_type: HA
platform: vsphere
openshift_version: 4.18.2
base_domain: vmware.example.com
cpu_architecture: x86_64

vSphere Requirements:

  • vSphere 7.0+ or 8.0+
  • vCenter Server access
  • Resource pool with adequate resources
  • Network portgroup configured

Network Configuration:

networking_mode: DHCP (from vSphere network)
machine_network: 172.16.10.0/24 (vSphere portgroup)
api_vip: 172.16.10.100
ingress_vip: 172.16.10.101

Host Requirements (VMs):

  • 3 control plane VMs: 4 vCPU, 16 GB RAM, 120 GB disk
  • 3 worker VMs: 8 vCPU, 32 GB RAM, 500 GB disk

Use Case:

  • Existing VMware infrastructure
  • Enterprise virtualized environment
  • Need vSphere integration (storage, DRS)

Post-Installation:

  • Configure vSphere CSI driver for dynamic storage
  • Enable DRS anti-affinity rules
  • Configure VM backup policies

Installation Time: ~60-90 minutes


Example 5: HA Cluster on Oracle Cloud (OCI)

Scenario: Cloud deployment on Oracle Cloud Infrastructure

Configuration:

cluster_name: oci-prod
cluster_type: HA
platform: oci
openshift_version: 4.18.2
base_domain: cloud.example.com
cpu_architecture: x86_64

OCI Requirements:

  • OCI account with quotas
  • Virtual Cloud Network (VCN) configured
  • Subnets for control plane and workers
  • Internet gateway or NAT gateway

Network Configuration:

networking_mode: OCI-managed
vips: Not required (OCI load balancers handle this)
machine_network: 10.0.0.0/16 (VCN CIDR)

Host Requirements (OCI Instances):

  • 3 control plane: VM.Standard.E4.Flex (4 OCPU, 16 GB RAM)
  • 3 workers: VM.Standard.E4.Flex (8 OCPU, 32 GB RAM)

Use Case:

  • Cloud-native deployment
  • Oracle Cloud customers
  • Need cloud load balancers and storage

Advantages:

  • Automatic load balancer provisioning (no VIPs)
  • Cloud-native networking
  • OCI Block Volumes for storage

Installation Time: ~60 minutes


Example 6: Development Cluster (Minimal)

Scenario: Local development and testing

Configuration:

cluster_name: dev-local
cluster_type: SNO
platform: none
openshift_version: 4.18.2
base_domain: dev.local
cpu_architecture: x86_64

Host Requirements:

  • 1 VM or physical server
  • 8 CPU, 32 GB RAM, 120 GB disk

Network Configuration:

networking_mode: DHCP
vips: Not required

DNS (local hosts file or dnsmasq):

192.168.122.100  api.dev-local.dev.local
192.168.122.100  oauth-openshift.apps.dev-local.dev.local
192.168.122.100  console-openshift-console.apps.dev-local.dev.local
192.168.122.100  *.apps.dev-local.dev.local

Use Case:

  • Developer workstation
  • CI/CD testing
  • Learning OpenShift
  • Proof of concept

Installation Time: ~30-45 minutes


Example 7: Multi-Cluster Setup for Submariner

Scenario: Two clusters connected via Submariner for multi-cluster apps

Cluster A Configuration:

cluster_name: cluster-a
cluster_type: HA
platform: baremetal
openshift_version: 4.18.2
base_domain: site-a.example.com
cpu_architecture: x86_64

networking:
  cluster_network: 10.128.0.0/14
  service_network: 172.30.0.0/16
  machine_network: 192.168.10.0/24
  api_vip: 192.168.10.100
  ingress_vip: 192.168.10.101

Cluster B Configuration:

cluster_name: cluster-b
cluster_type: HA
platform: baremetal
openshift_version: 4.18.2
base_domain: site-b.example.com
cpu_architecture: x86_64

networking:
  cluster_network: 10.132.0.0/14  # NON-OVERLAPPING
  service_network: 172.31.0.0/16  # NON-OVERLAPPING
  machine_network: 192.168.20.0/24
  api_vip: 192.168.20.100
  ingress_vip: 192.168.20.101

Critical Requirements:

  • ⚠️ Cluster and service networks MUST NOT overlap
  • Gateway nodes with public IPs or VPN connectivity
  • IPsec or WireGuard tunnels between clusters

Use Case:

  • Multi-site applications
  • Disaster recovery
  • Geo-distributed workloads
  • Hybrid cloud deployments

Post-Installation:

  • Install Submariner operator on both clusters
  • Configure gateway nodes
  • Establish connectivity and test

Example 8: Edge Cluster with Advanced Networking

Scenario: Edge location with bonded network interfaces and VLANs

Configuration:

cluster_name: edge-factory-01
cluster_type: SNO
platform: none
openshift_version: 4.18.2
base_domain: factory.example.com
cpu_architecture: x86_64

Network Configuration (Advanced Static):

networking_mode: Static (Advanced)

interfaces:
  - type: ethernet
    name: eth0
    mac_address: "52:54:00:aa:bb:01"
    # No IP - used for bonding

  - type: ethernet
    name: eth1
    mac_address: "52:54:00:aa:bb:02"
    # No IP - used for bonding

  - type: bond
    name: bond0
    mode: active-backup
    ports: [eth0, eth1]
    # No IP - base for VLAN

  - type: vlan
    name: bond0.100
    vlan_id: 100
    base_interface: bond0
    ipv4_address: 10.100.1.10
    prefix: 24

dns:
  servers: ["10.100.1.1", "8.8.8.8"]

routes:
  - destination: 0.0.0.0/0
    next_hop: 10.100.1.1
    interface: bond0.100

Use Case:

  • High-availability networking (bonding)
  • VLAN segmentation
  • Production edge deployments
  • Network redundancy requirements

Installation Time: ~60 minutes (complex networking)


Example 9: Compact Cluster (3 Nodes)

Scenario: Small production cluster with control plane nodes as workers

Configuration:

cluster_name: compact-prod
cluster_type: HA
platform: baremetal
openshift_version: 4.18.2
base_domain: compact.example.com
cpu_architecture: x86_64

Host Requirements:

  • 3 nodes (control plane + worker combined)
  • Each node: 8 CPU, 32 GB RAM, 500 GB disk

Network Configuration:

networking_mode: DHCP
machine_network: 192.168.50.0/24
api_vip: 192.168.50.100
ingress_vip: 192.168.50.101

Role Assignment:

  • All 3 nodes assigned "master" role
  • Nodes automatically schedulable for workloads

Use Case:

  • Small production environments
  • Cost-sensitive deployments
  • Limited hardware availability
  • Branch office deployments

Limitations:

  • Cannot scale workers separately
  • Higher load on control plane nodes

Installation Time: ~60 minutes


Example 10: Air-Gapped SNO

Scenario: Disconnected environment with no internet access

Configuration:

cluster_name: airgap-sno
cluster_type: SNO
platform: none
openshift_version: 4.18.2
base_domain: airgap.local
cpu_architecture: x86_64

Prerequisites:

  • Mirrored container registry (with all OpenShift images)
  • Mirrored RHEL package repository
  • Local NTP server
  • Local DNS server

Network Configuration:

networking_mode: Static
interface: eth0
mac_address: "52:54:00:aa:bb:01"
ipv4_address: 192.168.100.10
prefix: 24
gateway: 192.168.100.1
dns_servers: ["192.168.100.5"]  # Local DNS only

Additional Configuration:

image_content_sources:
  - mirrors:
    - registry.airgap.local:5000/openshift/release-images
    source: quay.io/openshift-release-dev/ocp-release
additional_trust_bundle: |
  -----BEGIN CERTIFICATE-----
  <registry-ca-cert>
  -----END CERTIFICATE-----

Use Case:

  • Classified/secure environments
  • No internet connectivity
  • Regulatory compliance (data sovereignty)

Installation Time: ~60 minutes (after mirroring complete)


Configuration Matrix

Example Type Nodes Platform Networking VIPs Use Case
1 SNO 1 none DHCP No Edge
2 HA 6 baremetal DHCP Yes Data center
3 HA 5 baremetal Static Yes Production
4 HA 6 vsphere DHCP Yes VMware
5 HA 6 oci OCI-managed No Cloud
6 SNO 1 none DHCP No Development
7 HA + HA 10 baremetal DHCP Yes Multi-cluster
8 SNO 1 none Static Advanced No Edge Advanced
9 Compact 3 baremetal DHCP Yes Small production
10 SNO 1 none Static No Air-gapped

Common Configuration Patterns

Pattern: Production HA

3 control plane nodes (dedicated)
3+ worker nodes (applications)
Static IPs or DHCP reservations
VIPs configured
Load balancer (external or VIPs)

Pattern: Edge SNO

1 node (all roles)
DHCP or static IP
No VIPs
Minimal resources (8 CPU, 32 GB RAM)

Pattern: Cloud-Native

HA cluster
Cloud platform (OCI, AWS, Azure)
Cloud load balancers (no VIPs)
Cloud storage integration

Pattern: Air-Gapped

Mirrored registry
Local DNS and NTP
Static networking
No internet access

References