This walkthrough describes how to configure a PXE server to load Windows PE by booting a client computer from the network. Using the Windows PE tools and a Windows 10 image file, you can install Windows 10 from the network.
A DHCP server: A DHCP server or DHCP proxy configured to respond to PXE client requests is required.
A PXE server: A server running the TFTP service that can host Windows PE boot files that the client will download.
A file server: A server hosting a network file share.
All four of the roles specified above can be hosted on the same computer or each can be on a separate computer.
Step 1: Copy Windows PE source files
On the deployment computer, select Start, and type deployment.
Right-click Deployment and Imaging Tools Environment and then select Run as administrator. The Deployment and Imaging Tools Environment shortcut opens a Command Prompt window and automatically sets environment variables to point to all the necessary tools.
Run the following command to copy the base Windows PE files into a new folder. The script requires two arguments: hardware architecture and destination location. The value of <architecture> can be x86, amd64, or arm and <destination> is a path to a local directory. If the directory doesn't already exist, it will be created.
Windows Command Prompt
copype.cmd <architecture> <destination>
For example, the following command copies amd64 architecture files to the C:\winpe_amd64 directory:
Windows Command Prompt
copype.cmd amd64 C:\winpe_amd64
The script creates the destination directory structure and copies all the necessary files for that architecture. In the previous example, the following directories are created:
Mount the base Windows PE image (winpe.wim) to the \mount directory using the DISM tool. Mounting an image file unpacks the file contents into a folder so that you can make changes directly or by using tools such as DISM. See the following example.
Verify that the message The operation completed successfully is displayed.
Note
To view currently mounted images, enter dism.exe /get-MountedWiminfo.
Map a network share to the root TFTP directory on the PXE/TFTP server and create a \Boot folder. Consult your TFTP server documentation to determine the root TFTP server directory, then enable sharing for this directory, and verify it can be accessed on the network. In the following example, the PXE server name is PXE-1 and the TFTP root directory is shared using a network path of \\PXE-1\TFTPRoot:
Windows Command Prompt
net.exe use y: \\PXE-1\TFTPRoot
y:
mdBoot
Copy the PXE boot files from the mounted directory to the \boot folder. For example:
Your PXE/TFTP server is now configured. You can view the BCD settings that have been configured using the command bcdedit.exe /store <BCD file location> /enum all. See the following example. Note: Your GUID will be different than the one shown below.
If you start the PXE boot process, but receive the error The boot configuration data for your PC is missing or contains error, then verify that \boot directory is installed under the correct TFTP server root directory. In the example used here the name of this directory is TFTPRoot, but your TFTP server might be different.
PXE boot process summary
The following process summarizes the PXE client boot.
Note
The following assumes that the client and PXE server are on the same network/subnet/vlan or that PXE requests have been appropriately forwarded from the client to the PXE server using IP helpers configured in the router or switch. For more information about IP helpers, see Configuring Your Router to Forward Broadcasts.
A client contacts the PXE server. When the client is on a different network/subnet/vlan as the PXE server, the client is routed to the PXE server using the IP helpers.
The PXE server sends DHCP options 060 (client identifier PXEClient), 066 (boot server host name) and 067 (boot file name) to the client.
The client downloads boot\PXEboot.n12 from the TFTP server based on DHCP option 067 boot file name value received from the PXE server.
PXEboot.n12 immediately begins a network boot.
The client downloads boot\bootmgr.exe and the boot\BCD file from the TFTP server.
Note
The BCD store must reside in the \boot directory on the TFTP server and must be named BCD.
Bootmgr.exe reads the BCD operating system entries and downloads boot\boot.sdi and the Windows PE image (boot\boot.wim). Optional files that can also be downloaded include TrueType fonts (boot\Fonts\wgl4_boot.ttf) and the hibernation state file (\hiberfil.sys) if these files are present.
Bootmgr.exe starts Windows PE by calling winload.exe within the Windows PE image.
Windows PE loads, a command prompt opens and wpeinit.exe is run to initialize Windows PE.
The Windows PE client provides access to tools like imagex.exe, diskpart.exe, and bcdboot.exe using the Windows PE command prompt. With the help of these tools accompanied by a Windows 10 image file, the destination computer can be formatted properly to load a full Windows 10 operating system.
Plan and execute an endpoint deployment strategy, using essential elements of modern management, co-management approaches, and Microsoft Intune integration.