Skip to content
Branch: master
Find file History
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

About Scenario

Description and requirements

  • In this scenario 2-16 node S2D cluster can be created from scratch using SCVMM

  • It is just simulation "how it would look like". Performance is not a subject here - it is just to test look and feel

  • Script is well tested, on both real and simulated environments. However you should ask your Premier Field Engineer to adapt scenario to your infrastructure

  • Nano Servers for bare metal deployment are not the best. It's deployed successfully, but vSwitch fails to apply for some reason.

  • Windows 10 1511 with enabled Hyper-V or Windows 10 1607+ (if nested virtualization is enabled)

  • 10GB Memory or 20GB if nested virtualization is used (for 4 node configuration)

  • SSD is a must

  • Please hydrate your main lab with SCVMM as demonstrated in this video

  • you can download SCVMM 1801 from eval center

  • not all best practices are in here. All best practices are in S2D Hyperconverged Scenario


$LabConfig=@{ DomainAdminName='LabAdmin'; AdminPassword='LS1setup!'; Prefix = 'WSLabSCVMM-'; SwitchName = 'LabSwitch'; DCEdition='ServerDataCenter'; VMs=@();InstallSCVMM='Yes' }

#these 2 VMs are not needed, if you are pasting scripts to DC and if WDS is on DC
$LabConfig.VMs += @{ VMName = 'Management' ; Configuration = 'Simple' ; ParentVHD = 'Win10_G2.vhdx'  ; MemoryStartupBytes= 1GB ; MemoryMinimumBytes=1GB ; AddToolsVHD=$True ; DisableWCF=$True }
$LabConfig.VMs += @{ VMName = 'WDS' ; Configuration = 'Simple' ; ParentVHD = 'Win2016_G2.vhdx'  ; MemoryStartupBytes= 1GB ; MemoryMinimumBytes=512MB  }


Tip: You can hydrate lab faster, if you already have parent disks created. Just copy ParentDisks to your folder together with scripts, and 2_CreateParentDisks.ps1 will skip parent disks creation, and will even create DC with SCVMM just by reusing parentdisk (if you have GUI version there). Best practice would be to hydrate all with core and use Win10 for management. Both scenarios are valid. I like Win10 more as it's the way it should be in production environment.

Tip 2: Copy Server Core VHD into tools.vhd as you will use it for bare metal deploy

Fun fact You need GUI version of Windows Server since WDS is in full version only.

Since script is bit long, all regions are described below without code snippets. All code is in scenario.ps1

region check prerequisites

This region just checks if VMM console is installed and also if RSAT components are present. It is valid for Server/ServerCore/Client. If VMM console is not installed, it will ask for setup.exe and install it for you.

region Variables

Here are all variables, that you can change - like domain name, networks, vSwitch name...

Script will ask you for credentials. Just provide your LabAdmin creds (it's for run as admin creds.,typically corp\LabAdmin LS1setup!). In real environment you would have dedicated account.

Script will also ask you for VHD. Just provide Core server VHD from your ParentDisks (copy it to tools disk before hydration as instructed in Tip 2)

region basic SCVMM Configuration

For some reason VMM service does not start, so its started if not running. Also some basic settings are configured like HostGroup, RunAsAccounts...

region Configure networks

region Configure Virtual Switch

region Configure Physical Computer Profile

region Configure WDS

Note: following warnings are expected

region Run from Hyper-V Host to create new VMs

This piece needs to run from Hyper-V host. And also you need to mofify VMPath variable to reflect your lab (in my case E:drive)

VMs will start and attempt to boot. This is crucial as we will be able to grab info about VMs from log.

region Deploy hosts (run again from DC or management machine)

This region can be easily modified to use it with real hardware. In virtual environment, it will build hash table with MAC addresses, IPs and names from SCVMM log.

To initiate deployment, just reboot S2D VMs on host after script will create deployment jobs

region Apply vSwitch

can take forever. Just wait. And validate all was deployed OK as on following screenshot. It sometimes does not apply vSwitch on first node.

region Configure Networking (classic approach)

In this part will script configure best practices (pNIC to vNIC mapping, static IP addresses, DCB ...). All steps are done with native (nonVMM) commandlets.

region Configure Cluster and S2D (classic approach)

In this region will be Cluster configured (cluster creation, cluster networks config) and will be S2D enabled. All usin native commandlets.

For all best practices review S2D Hyperconverged scenario .

region Create some Volumes (classic approach)

Here will be some volumes created using native PowerShell commands, and then renamed.

region Create some dummy VMs (3 per each CSV disk)

region add storage provider to VMM

In this part we will just add pool and volume management to VMM

You can’t perform that action at this time.