Windows Server Failover Clustering

# Windows Server Failover Clustering

Introduction

Windows Server Failover Clustering (WSFC) provides high availability for server roles and applications by grouping multiple servers into a cluster. If one node fails, another node takes over the workload with minimal downtime. This tutorial covers cluster creation, validation, quorum configuration, shared storage, cluster roles, and failover testing.

By the end of this guide, you will be able to deploy and manage a Windows Server Failover Cluster for high availability.

Technical Background

Failover Clustering Concepts

| Concept | Description | |---------|-------------| | Cluster | Group of servers working together for high availability | | Node | Individual server in the cluster | | Cluster Role | Service or application managed by the cluster | | Quorum | Voting mechanism to prevent split-brain | | Shared Storage | Storage accessible by all cluster nodes | | Failover | Automatic movement of role to another node | | Failback | Movement of role back to original node |

Quorum Modes

| Mode | Description | Node Count | |------|-------------|------------| | Node Majority | Each node votes; majority wins | Odd number of nodes | | Node and Disk Majority | Nodes + shared disk witness | Even number with disk | | Node and File Share Majority | Nodes + file share witness | Even number with file share | | No Majority: Disk Only | Disk witness determines quorum | Legacy, not recommended |

Prerequisites

• 2 or more Windows Server 2025 nodes
• Static IP addresses on all nodes
• Shared storage (iSCSI, FC, or SMB) accessible by all nodes
• Network connectivity between nodes
• All nodes in same domain
• Failover Clustering feature installed

Architecture

Two-Node Cluster with Disk Witness

Node 1                          Node 2
┌─────────────┐                 ┌─────────────┐
│   WSFC      │                 │   WSFC      │
│   Node 1    │                 │   Node 2    │
└──────┬──────┘                 └──────┬──────┘
       │                              │
       └──────────┬───────────────────┘
                  │
        ┌─────────▼─────────┐
        │   Shared Storage  │◄── Disk Witness
        │   (iSCSI/FC/SMB)  │
        └───────────────────┘

Multi-Site Cluster

Site A                          Site B
┌─────────┐                     ┌─────────┐
│  Node 1 │──── WAN/VPN ───────▶│  Node 3 │
│  Node 2 │                     │  Node 4 │
└────┬────┘                     └────┬────┘
     │                              │
     └────────┬─────────────────────┘
              │
        ┌─────▼─────┐
        │File Share │◄── File Share Witness
        │   Witness │
        └───────────┘

Step-by-Step Implementation

Step 1: Install Failover Clustering Feature

**Purpose**: Install Failover Clustering feature on all cluster nodes.

**Commands** (PowerShell):

# Install Failover Clustering on each node
Install-WindowsFeature -Name Failover-Clustering -IncludeManagementTools

# Verify installation
Get-WindowsFeature Failover-Clustering

# Install additional tools (optional)
Install-WindowsFeature -Name RSAT-Clustering-Mgmt

**Expected output**: Failover Clustering feature installed on all nodes.

**Verification**:

Get-WindowsFeature Failover-Clustering | Select-Object Name, InstallState
# Shows: Installed

Step 2: Validate Cluster Configuration

**Purpose**: Run cluster validation to verify hardware, network, and storage compatibility.

**Commands** (PowerShell):

# Run cluster validation
Test-Cluster -Node Node1, Node2 -Include "Inventory", "Network", "Storage"

# View validation report
Get-ClusterReport -Path "C:\Windows\Cluster\Reports\Validation Report*.html" -Latest

# Or run from Server Manager: Failover Cluster Manager → Validate Configuration

**Expected output**: Validation report shows all tests passed or warnings addressed.

**Verification**:

# Review validation report HTML
# All critical tests should pass

Step 3: Configure Shared Storage

**Purpose**: Set up shared storage accessible by all cluster nodes.

**Commands** (PowerShell) - iSCSI example:

# On each node, discover and connect to iSCSI target
New-IscsiTargetPortal -TargetPortalAddress 192.168.1.60
Connect-IscsiTarget -TargetName "iqn.2024-01.com.storage:target1" -IsPersistent $true

# Verify disk visible on all nodes
Get-Disk | Where-Object {$_.IsSystem -eq $false}

# Initialize and format disk (on one node only)
Initialize-Disk -Number 1 -PartitionStyle GPT
New-Partition -DiskNumber 1 -UseMaximumSize -DriveLetter S
Format-Volume -DriveLetter S -FileSystem NTFS -NewFileSystemLabel "ClusterShared"

**Expected output**: Shared disk visible on all nodes with same drive letter.

**Verification**:

# On each node
Get-Disk -Number 1 | Select-Object Number, FriendlyName, HealthStatus, OperationalStatus

Step 4: Create Failover Cluster

**Purpose**: Create the cluster with validated nodes.

**Commands** (PowerShell):

# Create cluster (run from one node)
New-Cluster -Name "ITYANTRA-CLUSTER" -Node Node1, Node2 -StaticAddress 192.168.1.100 -NoStorage

# Verify cluster creation
Get-Cluster
Test-Cluster -Node Node1, Node2

# Add shared storage to cluster
Add-ClusterDisk -Disk "Cluster Shared Disk" -Cluster ITYANTRA-CLUSTER

# Verify cluster nodes and storage
Get-ClusterNode
Get-ClusterAvailableDisk
Get-ClusterResource

**Expected output**: Cluster created with nodes and shared storage.

**Verification**:

Get-Cluster
# Shows cluster name, nodes, and status

Step 5: Configure Quorum

**Purpose**: Set quorum mode for cluster decision-making.

**Commands** (PowerShell):

# Set quorum mode (Node Majority for odd number of nodes)
Set-ClusterQuorum -NodeMajority

# Or Node and Disk Majority (even number of nodes with disk witness)
Set-ClusterQuorum -NodeAndDiskMajority -Disk "Cluster Disk 1"

# Or Node and File Share Majority
Set-ClusterQuorum -NodeAndFileShareMajority "\\FileServer\QuorumShare"

# Verify quorum configuration
Get-ClusterQuorum

**Expected output**: Quorum mode configured appropriately.

**Verification**:

Get-ClusterQuorum
# Shows quorum mode and witness

Step 6: Configure Cluster Role (Highly Available Service)

**Purpose**: Make a service or application highly available.

**Commands** (PowerShell):

# Add File Server role
Add-ClusterFileServerRole -Name "FS01" -Storage "Cluster Disk 1"

# Add Generic Application role
Add-ClusterGenericApplicationRole -Name "App01" -CommandLine "C:\App\app.exe" -Storage "Cluster Disk 1"

# Add Generic Service role
Add-ClusterGenericServiceRole -Name "Svc01" -Service "MyService" -Storage "Cluster Disk 1"

# Verify cluster roles
Get-ClusterRole

**Expected output**: Cluster role created and running on one node.

**Verification**:

Get-ClusterRole
# Shows role name, state, owner node

Step 7: Test Failover

**Purpose**: Verify failover works correctly.

**Commands** (PowerShell):

# Move cluster role to another node
Move-ClusterGroup -Name "FS01" -Node Node2

# Verify role moved
Get-ClusterRole -Name "FS01"

# Simulate node failure (stop cluster service on Node1)
Stop-Service -Name ClusSvc -ComputerName Node1 -Force

# Verify role failed over to Node2
Get-ClusterRole -Name "FS01" | Select-Object Name, State, OwnerNode

# Restart Node1 cluster service
Start-Service -Name ClusSvc -ComputerName Node1

**Expected output**: Role fails over to another node without data loss.

**Verification**:

Get-ClusterRole -Name "FS01" | Select-Object Name, State, OwnerNode
# Shows OwnerNode changed to Node2

Configuration Examples

Cluster Validation Command

Test-Cluster -Node Node1, Node2 -Include "Inventory", "Network", "Storage"

Cluster Creation with Static IP

New-Cluster -Name "ITYANTRA-CLUSTER" -Node Node1, Node2 -StaticAddress 192.168.1.100

Add Cluster Disk

Get-ClusterAvailableDisk | Add-ClusterDisk

Configure Cluster Network

# Set cluster network priority
(Get-ClusterNetwork "Cluster Network 1").Metric = 100
(Get-ClusterNetwork "Cluster Network 2").Metric = 200

# Enable/disable cluster network
(Get-ClusterNetwork "Cluster Network 1").Role = 3  # 1=Enabled, 2=Disabled, 3=Enabled for cluster only

Cluster Role Properties

# Set role failback behavior
(Get-ClusterRole "FS01").FailbackType = 1  # 0=Never, 1=Allow, 2=Prevent

# Set failover threshold
(Get-ClusterGroup "FS01").FailoverThreshold = 2
(Get-ClusterGroup "FS01").FailoverPeriod = 6  # hours

Verification

Cluster Verification Commands

Get-Cluster                 -- Cluster overview
Get-ClusterNode             -- Cluster nodes and status
Get-ClusterResource         -- Cluster resources
Get-ClusterRole             -- Cluster roles
Get-ClusterGroup            -- Cluster groups
Get-ClusterQuorum           -- Quorum configuration
Test-Cluster                -- Run validation tests
Get-ClusterLog              -- View cluster logs

Expected Results

• Cluster shows all nodes in "Up" state
• Cluster role shows "Online" status
• Shared storage visible to all nodes
• Quorum configured with appropriate mode
• Failover moves role to another node successfully

Troubleshooting

| Symptom | Possible Cause | Diagnostic Command | Corrective Action | Verification | |---------|---------------|-------------------|-------------------|--------------| | Cluster creation fails | Network or storage issue | `Test-Cluster` | Fix validation errors | Cluster created | | Node not in cluster | Cluster service not running | `Get-Service ClusSvc` | Start cluster service | Node joins | | Resource offline | Dependency missing | `Get-ClusterResource` | Fix dependencies | Resource online | | Failover fails | Incompatible node or missing storage | `Get-ClusterLog` | Add storage; validate nodes | Failover succeeds | | Quorum lost | Too many nodes down | `Get-ClusterQuorum` | Restore nodes; force quorum if needed | Quorum regained | | Split-brain | Network partition | Cluster logs | Restore network; force quorum | Single cluster | | Shared storage not visible | Zoning or LUN masking | `Get-Disk` | Fix storage zoning | Disk visible |

Best Practices

1. **Use odd number of nodes**: 3 nodes for Node Majority quorum (avoid tie votes). 2. **Use dedicated cluster network**: Separate cluster heartbeat from production traffic. 3. **Test failover regularly**: Verify failover works during maintenance windows. 4. **Configure appropriate quorum**: Match quorum mode to node count and storage availability. 5. **Use shared storage**: All nodes must access same storage for cluster roles. 6. **Monitor cluster health**: Alert on node failures and resource issues. 7. **Document cluster configuration**: Record nodes, storage, quorum, and roles. 8. **Keep nodes updated**: Apply same patches to all nodes. 9. **Use cluster-aware updating**: Update nodes without downtime. 10. **Test disaster recovery**: Verify cluster recovery after site failure.

Conclusion

Windows Server Failover Clustering provides high availability for critical server roles and applications. Successful cluster deployment requires validated hardware, shared storage, appropriate quorum configuration, and regular failover testing. With WSFC, applications remain available even during individual node failures, minimizing downtime and maintaining business continuity.

References

• Windows Server Failover Clustering Documentation (https://learn.microsoft.com/windows-server/failover-clustering/)
• PowerShell FailoverClustering Module (https://learn.microsoft.com/powershell/module/failoverclustering/)
• Cluster Quorum Modes (https://learn.microsoft.com/windows-server/failover-clustering/manage-cluster-quorum)
• Windows Server Clustering TechNet (https://learn.microsoft.com/windows-server/failover-clustering/)