[BUG] unable to create cluster with MetalLB - InvalidDiskCapacity?

**Describe the bug**
I try to deploy MetalLB as load balancer instead of the built-in HAproxy solution.
When i try to deploy a new cluster without a load balancer, the control-plane is getting stuck in deployment phase.

**To Reproduce**
Steps to reproduce the behavior:

Follow the guide https://docs.microsoft.com/en-us/azure-stack/aks-hci/deploy-metallb

When I execute **New-AksHciCluster -Name testcluster -nodePoolName testpool -nodeCount 2 -OSType linux -nodeVmSize Standard_A4_v2 -loadBalancerSettings $lbCfg**

I get the following message in powershell

**VERBOSE: [07/28/2022 15:45:54] [Kva] 2022-07-28T15:45:54+02:00 - INFO - WaitingForInfrastructure @ Machine/testcluster-control-plane-d6c8f: 1 of 2 completed**
....
**VERBOSE: [07/28/2022 15:10:48] [Kva] 2022-07-28T15:10:48+02:00 - INFO - KubeadmControlPlaneMachinesNotReady: Machines not in Running phase: **

**I never get passed this phase, I get the same message over and over again.**

**I get the following output from Get-AksHciCluster**

Status : {ProvisioningState, Details}
ProvisioningState : Deployed
KubernetesVersion : v1.22.6
PackageVersion : v1.22.6-kvapkg.1
NodePools : akspool01t
WindowsNodeCount : 0
LinuxNodeCount : 3
ControlPlaneNodeCount : 3
ControlPlaneVmSize : Standard_A4_v2
AutoScalerEnabled : False
AutoScalerProfile :
Name : akstest01t

Status : {ProvisioningState, Details}
ProvisioningState : ScalingControlPlane
KubernetesVersion : v1.22.4
PackageVersion : v1.22.4-kvapkg.2
NodePools :
WindowsNodeCount : 0
LinuxNodeCount : 0
ControlPlaneNodeCount : 1
ControlPlaneVmSize : Standard_A4_v2
AutoScalerEnabled : False
AutoScalerProfile :
Name: testcluster

**It seems that a node pool doesn't get created when I create the new cluster**

**Output from kubectl describe nodes from the newly created cluster.**

Events:
Type Reason Age From Message
Normal Starting 72s kube-proxy
Warning InvalidDiskCapacity 82s kubelet invalid capacity 0 on image filesystem
Normal NodeHasSufficientMemory 82s kubelet Node moc-lcts4zypxgg status is now: NodeHasSufficientMemory
Normal NodeHasNoDiskPressure 82s kubelet Node moc-lcts4zypxgg status is now: NodeHasNoDiskPressure
Normal NodeHasSufficientPID 82s kubelet Node moc-lcts4zypxgg status is now: NodeHasSufficientPID
Normal NodeAllocatableEnforced 82s kubelet Updated Node Allocatable limit across pods
Normal Starting 82s kubelet Starting kubelet

This seems to be the issue, that the nodepool doesn't get created and hence the control-plane doesn't have any storage to store its filesystem?
** Warning InvalidDiskCapacity 82s kubelet invalid capacity 0 on image filesystem **

**Expected behavior**
Deploy a new cluster and apply MetalLB as loadbalancer.

**Environment (please complete the following information):**
- OS Name: Microsoft Azure Stack HCI
OS Version: 10.0.20348 N/A Build 20348]
- Kubernetes Version (1.22.4 & 1.22.6 somehow?)

**Collect log files**