Hello Robert, I am Henry and I want to share my insight about your issue.
My recommendation is to build the cluster in a way that bypasses the S2D checks from the start. You can refer the workflow to configure your Failover Cluster correctly.
Step 1: Prerequisites - Before creating the cluster, ensure the storage is correctly prepared on all nodes.
- Install MPIO: On all server nodes, install the Multi-Path I/O (MPIO) feature. This is critical for redundancy and performance with shared block storage like Oracle PCA. A reboot will be required.
- Present Storage: Ensure the shared LUNs from your Oracle PCA are presented uniformly to all server nodes.
- Initialize Disks: On one node, use Disk Management to bring the disks Online and Initialize them as GPT.
- Crucial: Do not format the disks or assign them drive letters. They should remain as unallocated space. On all other nodes, the disks should simply appear as Offline.
Step 2: Run Cluster Validation (Ignoring S2D)
On one of the nodes, open PowerShell as an Administrator.
- Run the validation test, specifically telling it to ignore the S2D checks that are causing the errors:
Test-Cluster -Node <Node1,Node2> -Ignore ``'Storage Spaces Direct'
- Review the validation report. Resolve any reported errors related to your network or storage configuration before proceeding.
Step 3: Create the Cluster (Without Adding Storage)
The key to bypassing S2D during creation is the -NoStorage parameter.
- In the same administrative PowerShell window, run the following command:
New-Cluster -Name ``<YourClusterName>`` -Node <Node1,Node2> -StaticAddress ``<YourClusterIP>`` -NoStorage
- This creates a functional cluster object without attempting to automatically claim storage, thus avoiding the S2D logic entirely.
Step 4: Add and Configure Storage
- Now that the cluster exists, you can manually add your prepared shared disks.
- Open Failover Cluster Manager.
- Navigate to Storage > Disks and click Add Disk in the Actions pane. Your Oracle PCA LUNs will be discovered and added to "Available Storage."
- Configure Quorum: Right-click a smaller disk (e.g., 1-5 GB) and select "Configure Cluster Quorum Settings" to use it as a Disk Witness. This is critical for cluster stability.
- Create Cluster Shared Volumes (CSV): For the larger data disks that will host your services, right-click them and select "Add to Cluster Shared Volumes". This makes the storage accessible to all nodes simultaneously under C:\ClusterStorage.
Step 5: Test High Availability
With the cluster and storage now properly configured, test the failover functionality.
- Create a test role on the cluster (e.g., a test virtual machine or a simple "File Server" role) and place its resources on one of your new CSVs.
- In Failover Cluster Manager, right-click the test role, select Move > Best Possible Node.
- Confirm that the role and its storage successfully move to the other node and come online without errors.
I hope this information is helpful.