How to Create a Windows server 2022 SOFS Cluster on a VM #sofs #ws2022 #winserv #hyperv #Azure #Windows11 #WiMVP

Building a test lab is always depending on the resource you have. Building a Lab in Azure is giving you unlimited resources and the method on building this on your own laptop. I will use the GUI as much as possible else with a powershell script there is no fun in writing the blog. I’ll use a Windows 11 OS for this blog.

For now this blog will demonstrate how to create a scale out fileserver on a windows server 2022 platform.

First we have two domain member servers ws2022 01 / 02 .

Installing the Cluster Roles on the server with powershell or the GUI

Get-WindowsFeature Failover-Clustering
install-WindowsFeature “FS-FileServer”,”Failover-Clustering”,”RSAT-Clustering” -IncludeAllSubFeature -IncludeManagementTools -ComputerName “ws2022-01.mvp.local”,”ws2022-02.mvp.local”
#Create cluster validation report
Test-Cluster -Node ws2022-01,ws2022-02
Start-sleep 8
New-Cluster -Name ws2022CL01 -Node ws2022-01,ws2022-02 -NoStorage -StaticAddress “”

Remember installing the fileserver may take a reboot, you can also do this at a later stage to avoid the extra reboot.

Now that the cluster is created we configure the cluster, Quorum and add storage to the Cluster.

Here is the difference between the local setup and an Azure setup or running on windows 11. Personally I run Windows server as desktop.

Adding storage to the VM is done in the hyper-v manager. If you run Server! If you run Windows 10 or 11 you will face the issue explained below

Make sure you use scsi disk and shared disks else the disks are unusable for the SOFS file cluster. First option is create 3 shared disks

Make sure you using a shared location to store the vhd files.

When using Windows server you can bypass the share location by using a filterdriver fltMC.exe attach svhdxflt I:\ this is not working on windows 11 it is part of the Failover Clustering feature and will only work on Windows Server!

fltmc.exe attach svhdxflt C:\

Attach failed with error: 0x801f0013
The system could not find the filter specified.

To by pass this you can use ISCSI on the VM’s and this can also work perfectly on Azure. As it is a test lab the performance maybe a bit less of the iscsi connection, but works just as good.

So for the shared disk I create 3 iscsi targets each disk is mounted to both VM’s with the build in iscsi initiator. make sure the disks are not formatted and online.

Checking our just created Cluster on ws2022, and make sure it you work on node 1 all the resources are also available on node 1 , not that the disks are sitting on node 2. you could also pause node 2 that way you make sure there are no resources running on that node.

Add disk if you want a normal file server, but we are building a SOFS with CA storage, as I don’t want to wait if the disk is failing over. as a file server is way different that a SOFS!

In this case we want to build a scale out file server so we are not adding the disk here but we going create a disk pool.

A new pool is created , next step is a virtual disk and a volume

When there are no disk available the cluster is not visible here.

a minimum of 3 disks are needed, and in you test lab it can be any size but bigger that 16Gb

creating the pool,

now that the pool is created, we create the disk

The new disk is created in the next step new virtual disk

as we only have 3 disks and two nodes we have limited of configuration options.

I go for a Mirror as this will max my performance, the more disk you have the more performance you will get and different type of disk can also give caching if needed, with modern hardware Gb/s speed is easy done.

I choose here 50Gb but it all depends on the need and disk size you have. I have 1,49TB but i want to create more disk later so i need some space. and I have zero workload here.

When completed we have a virtual disk and just need to create a volume. I did uncheck the box as adding the volume on a different method, same result but just showing you that the cluster is interacting with the file server components.

When created there is a checkbox checked for the blog I unchecked this. Now I have created a Pool With a disk and our last step is creating a volume on that vdisk.

Now that the pool,disk,volume is created we can create the SOFS, must say the SOFS can be created first and add the disk later. but I like to do this this way.

Create the Scale out file server

Make sure you choose Scale out file server, the default is file server.

This will also be your netbios name. Can be changed but better use the correct name. It will be a Distributed network name.

As our final step we add the file share and this share is on top of our CSV volume that we created on the disk pool.

Add a fileshare

Just pick quick

Make sure the disk is also on your connected node, if not then you will not see the cluster storage

Create a share name.

Make sure the checkbox is set on Continuous Availability

Make sure you set the access rights conform your needs.

Make sure you set the permissions right on the file share. and grant the cluster node access to the share.

Then there comes the fun part testing performance

As you can see there is a nice performance on my test lab machine on a 1 core VM.

See good perfromance on just to see how things are working or giving a good demo

Hope it was helpfull thanks for visiting my blog.

Follow Me on Twitter @ClusterMVP

Follow My blog

Linkedin Profile Http://