WDS: Performing Multicast Deployments

Applies To: Windows Server 2008, Windows Server 2008 R2

In order to deploy an image using multicasting instead of unicasting, you must first create a multicast transmission. Multicast transmissions make the image available for multicasting, which enables you to deploy an image to a large number of client computers without overburdening the network. When you deploy an image using multicasting, the image is sent over the network only once, which can drastically reduce the amount of network bandwidth that is used. Note that data is transferred only if clients request data. If no clients are connected (that is, if the transmission is idle), data will not be sent over the network. Note that this topic pertains only to the default installation of Windows Deployment Services. If you only installed the Transport Server role service, see one of the following for instructions:

• Using Transport Server if you have Windows Server 2008
  
• Configuring Transport Server if you have Windows Server 2008 R2
  

In this topic

• When to Implement Multicasting
  
• Prerequisites for creating a multicast transmission
  
• Known issues in creating a multicast transmission
  
• Steps for creating a multicast transmission
  
• Additional Tasks
  

When to implement multicasting

 

Consider implementing multicasting if your organization:

Multicasting might not optimize your installations if your organization:

Has network routers that support multicasting. Is a large company that requires many concurrent client installations. Wants to use network bandwidth efficiently. This is because with this feature, images are sent over the network only once, and you can specify limitations (for example, to only use 10 percent of your bandwidth). Has enough disk space on client computers for the image to be downloaded. Meets the requirements listed in the following section.

Has network routers that do not support multicasting. Does not have bandwidth overload problems. Deploys images to only a small number of client computers simultaneously. Has disk space limitations on the client computers. (This is because the image is downloaded to client computers instead of being installed from a server.)

Prerequisites for creating a multicast transmission

• Routers that support multicasting. In particular, Internet Group Membership Protocol (IGMP) snooping should be enabled on all devices. This will cause your network hardware to forward multicast packets only to those devices that are requesting data. If IGMP snooping is turned off, multicast packets are treated as broadcast packets, and will be sent to every device in the subnet.
  
• At least one install image that you want to transmit on the server.
  
• The Boot.wim file from the product DVD for one of the following operating systems:
  
  
  • Client: Windows Vista (with at least Service Pack 1 (SP1)) or Windows 7. Do not use the Boot.wim from the Windows Vista DVD unless your version of Windows Vista has SP1 integrated into the DVD. If you use the Windows Vista Boot.wim, you will be able to create the transmission, but users who boot into it will not be able to join the transmission.
    
  • Server: Windows Server 2008 or Windows Server 2008 R2.
    
  You can also use a custom boot image. For instructions, see Creating Images (http://go.microsoft.com/fwlink/?LinkId=115311).
  

Known issues in creating a multicast transmission

You may encounter the following issues when implementing multicasting:

• If multiple servers are using multicast functionality on a network (Transport Server, Deployment Server, or another solution), it is important that each server is configured so that the multicast IP addresses do not collide. Otherwise, you may encounter excessive traffic when you enable multicasting. Note that each Windows Deployment Services server will have the same default range. To ensure that each server is using a unique IP address, use Multicast Address Dynamic Client Allocation Protocol (MADCAP) or specify static ranges that do not overlap. To specify a range, right-click the server in the MMC snap-in, click Properties, and then click the Network Settings tab.
  
• After you configure Windows Deployment Services server, if you modify the Multicast IP Address, the UDP port range, or the RPC port number, you must restart the service before the changes will take effect. If you do not restart the service, the server will use the old values and may not answer clients. To restart the service, right-click the server in the MMC snap-in, click All Tasks, and then click Restart.
  
• Each transmission can be run only as fast as the slowest client. That is, the entire transmission will be slow if there is one slow client. If you have Windows Server 2008 R2, you can resolve this issue by setting the transfer policy as specified in step 1 of the following procedure. Otherwise, to resolve this issue, first determine the client that is holding back the transmission (this is called the master client). To do this, open an elevated Command Prompt window and run WDSUTIL /Get-MulticastTransmission /Show-clients. Next, disconnect the master client. This will force the master client to run the transmission by using the Server Message Block (SMB) protocol, and the other clients' multicast performance should speed up. If they do not speed up, repeat this step for the new master client.
  

Steps for creating a multicast transmission

To perform this task, use the following procedure.

To enable multicasting of an install image

1. If you have Windows Server 2008 R2, you should first set the transfer policy for the server (if you have the initial release of Windows Server 2008, skip this step). To do this, right-click the server, click Properties, click the Multicast tab, and select an option under Transfer Settings. These settings allow you to enable the following:
   
   • Multiple stream transfer. The first option under Transfer Settings uses a single stream for all multicast clients, regardless of client speed. The next two options allow you to separate slower clients into their own multicast stream, which enables faster clients to complete their deployment more quickly instead of being held back by slower clients. These settings are only available for clients that boot into a boot image from Windows 7 or Windows Server 2008 R2.
     
   • Automatic disconnect. The last option under Transfer Settings allows you to automatically stop clients that fall under a specified speed. The clients that are disconnected will continue to transfer the image using unicasting (Server Message Block).
     
2. To create a multicast transmission for an install image, right-click the Multicast Transmissions node (or an install image) and click Create Multicast Transmission.
   
3. Type a name for the transmission.
   
4. Continue to follow the instructions in the wizard. There are two types of multicast transmissions:
   
   • Auto-Cast. This option indicates that as soon as an applicable client requests the image, a multicast transmission of the selected image begins. Then, as other clients request the same image, they too receive the image using multicasting.
     
   • Scheduled-Cast. This option sets the start criteria for the transmission based on the number of clients that are requesting an image and/or a specific day and time. With ScheduledCast, only clients that join before the transmission is started will receive the image using multicasting. That is, clients that join later will receive the image using unicasting. If you do not select start criteria for the transmission, then you must manually start it (by right-clicking the transmission and clicking Start).
     

    

   Consider using Auto-Cast if:

   Consider using Scheduled-Cast if:

   You work for a large corporation or an organization with high bandwidth that can handle installations at any time. You do not want customers to have to wait for the installation to begin.

   You work for a smaller organization or an organization where network traffic is an issue during the day. This way, you can set installations to occur during nonpeak hours or at night. To reduce the total time of the transmission. Because you can set multiple clients to start at the same time, the time will be reduced because Windows Deployment Services will not have to resend a part of the image to clients that started after the first client. You do not want the transmission to start until you manually start it (to do this, clear both check boxes when you create the transmission).

Additional Tasks

The following are additional tasks that you can perform to configure and manage multicast transmissions.

To configure transmissions

After a transmission has been created, you can perform the following actions:

• Start the transmission. If the transmission is the Scheduled-Cast type, there is at least one client, and the transmission has not started yet, you can right-click the transmission and then click Start. Alternatively, you can run WDSUTIL /Start-MulticastTransmission /Image: /ImageType:Install /ImageGroup:.
  
• Delete the transmission. If you right-click the transmission and click Delete, the multicast transmission stops and each client installation will start the transmission again using unicasting. That is, the client installations will not be deleted completely, but they will not use the multicast transmission to complete the installation. Alternatively, you can run WDSUTIL /Remove-MulticastTransmission /Image: /ImageType:Install /ImageGroup: /Force.
  
• Deactivate the transmission. If you right-click and then click Deactivate, each client that is currently installing will continue, but no new clients will be joined to the transmission. After each current client installation is completed, the transmission will be deleted. If there are no clients when you click this option, the transmission will be deleted instantly. Altnernatively you can run WDSUTIL /Remove-MulticastTransmission /Image: /ImageType:Install /ImageGroup:.
  
• View the transmission's properties. To view the properties, right-click the transmission and then click Properties. Note that you cannot edit the properties of a transmission after it is created. To make a change after you have created a transmission, you need to delete it and then recreate it. Alternatively, you can run WDSUTIL /Get-MulticastTransmission /Image: /ImageType:Install /ImageGroup:.
  
• Refresh the transmissions and data. To do this, right-click a transmission and then click Refresh. You can also refresh the data by pressing F5.
  

To configure clients in a transmission

After you have created the transmission, client computers can join it by selecting the image in the Windows Deployment Services client installation screens. Client computers can also join a transmission by using Wdsmcast.exe, a command-line tool included in the Windows Automated Installation Kit (Windows AIK). When there are clients in a transmission, you can do any of the following:

• View clients and see progress. To view any connected clients, expand the Multicast Transmissions node, click the image, and view the right pane. Windows Server 2008 R2 includes additional progress information including the master client and performance counters such as the client’s memory, CPU, and network usage. Alternatively, you can run WDSUTIL /Get-MulticastTransmission /Image: /ImageType:Install /ImageGroup: /show:clients or to view a client’s , run WDSUTIL /Get-MulticastTransmission /Image: /ImageType:Install /ImageGroup: /show:clients.
  
• Stop a client installation. To stop the installation completely, right-click a client and then click Disconnect. You should use this option with caution because the installation will fail and the computer could be left in an unusable state. Alternatively you can run WDSUTIL /Disconnect-Client /ClientID: /Force.
  
• Disconnect a client from a multicast transmission. To discontinue the transmission for a particular client but continue to transfer the image through unicasting, right-click the client, and then click Bypass multicast. WDSUTIL /Disconnect-Client /ClientID:.
  

To configure the UDP port range for multicasting

This setting specifies the range of UDP ports to use for multicasting and other components, such as the Trivial File Transfer Protocol (TFTP) provider. Before you change this range, you need to have at least as many ports as you have sessions and concurrent clients accessing the server. In terms of multicasting, a session is a network interface on your server. To calculate the number of sessions, multiply the number of network adapters on your server by the number of images that could be concurrently transferred using multicasting. For example, if you have two network adapters, and clients are connected on both interfaces, the content will be sent on the network twice (once from each interface). So in this case, you would need at least two ports. Because this range is also used by the TFTP provider, you will need as many available ports as you have concurrent clients accessing the server.

 

Using the MMC	Using WDSUTIL
Right-click the server, and then click Properties. On the Network Settings tab, specify the UDP port range. You must restart the service before the changes will take effect. To do this, right-click the server in the MMC snap-in, click All Tasks, and then click Restart.	Click Start, right-click Command Prompt, and click Run as administrator. Run WDSUTIL /Set-Server [/Server:] /Transport /StartPort:x /EndPort:y. You must restart the service before the changes will take effect. To do this, run wdsutil /stop-server and then run wdsutil /start-server.

To configure how the server will obtain IP addresses for multicasting

The server allocates a multicast IP address to each multicast session, and all connected clients listen in on that address. It's important that all IP addresses be unique on the network to ensure that each client receives the correct data. If you have a complex network, you should consider using DHCP to select the addresses. In more basic environments, you can configure a range and have the Windows Deployment Services server select the address.

 

Using the MMC

Using WDSUTIL

Right-click the server, and then click Properties. On the Network Settings tab (or the Multicast tab for Windows Server 2008 R2) under Multicast IP Address, select one of the following: Obtain IP address from DHCP Use IP address from the following range Right-click the server in the MMC snap-in, click All Tasks, and then click Restart.

Click Start, right-click Command Prompt, and click Run as administrator. Do one of the following: To use MADCAP to obtain the IP address for each namespace, run WDSUTIL /Set-Server [/Server:] /Transport /ObtainIPFrom:DHCP. To configure a preset range of IP addresses, run WDSUTIL /Set-Server [/Server:] /Transport /ObtainIPv4From:Range /Start:x.x.x.x /End:y.y.y.y. You must restart the service before the changes will take effect. To do this, run WDSUTIL /stop-server and then run WDSUTIL /start-server.

To avoid IP address conflicts

When two servers select the same multicast IP address to send data to, data intended for clients of either server can be routed to all clients. This causes unnecessary network traffic. Note also that this is particularly harmful if the servers are connected by a low-bandwidth connection (such as a wide area network (WAN) link), because both sets of data will be sent over this connection. The following are preventive measures that you should take to avoid this situation. To modify these options, right-click the server in the MMC snap-in, click Properties, and then click the Network Settings tab (or the Multicast tab for Windows Server 2008 R2).

• Use DHCP for IP addresses. This will prevent addresses from being assigned twice.
  
• Configure a static range for each server, making sure that this range does not overlap with the ranges defined for other servers.
  
• Lower the multicast Time-To-Live (TTL) setting to prevent the routers from forwarding multicast traffic outside the site network. You can also configure your border router not to forward multicast traffic.

WDS: Performing an Unattended Installation

Applies To: Windows Server 2008, Windows Server 2008 R2

You can automate the entire deployment for Windows Deployment Services. To do this, you will need to create two different unattend files (one for the Windows Deployment Services UI screens, and one for the later phases of Setup) and associate them using the procedures in this topic. Two files are necessary because Windows Deployment Services can deploy images that support the Unattend.xml format as well as images that do not support the Unattend.xml format (Windows Server 2003 and older versions of Windows).
This topic contains information about how to create these files, and instructions about how to configure the files on the server in order to automate the deployment.
For more detailed information, see Performing Unattended Installations (http://go.microsoft.com/fwlink/?LinkId=89226).

Prerequisites for performing an unattended installation

• Ensure that you are a member of the Local Administrators group on the Windows Deployment Services server.
  
• Ensure that you have sufficient credentials to add a computer to a domain. For more information, see Required Permissions (http://go.microsoft.com/fwlink/?LinkId=115301).
  

Windows Deployment Services client unattend file

First, create an Unattend.xml file with settings applicable to Windows Deployment Services client UI screens (for entering credentials, choosing an install image, configuring the disk, and so on). You should author the file by using Windows System Image Manager (SIM) from the Windows Automated Installation Kit (Windows AIK). For example files, see Sample Unattend Files (http://go.microsoft.com/fwlink/?LinkId=122642). For details about each section of the file, see the “Automating the User Interface Screens of the Windows Deployment Services client” section of Automating Setup (http://go.microsoft.com/fwlink/?LinkId=89226).
After you have created the file, you must associate it with a client. There are three ways you can do this:

• Associate the client unattend file for all architectures. You can set a default client unattend file for all booting clients based on the architecture of the client. To associate an unattend file, use the following procedure or run WDSUTIL /Set-Server /WDSUnattend /Policy:enabled /File:wdsclientunattend\unattend.xml /Architecture:.
  
  
  To associate a client unattend file by architecture

  1. Copy the client unattend file to RemoteInstall\WDSClientUnattend.
  2. Open the Windows Deployment Services MMC snap-in, right-click the server that contains the image that you want to associate the unattend file with, and then click Properties.
     
  3. On the Client tab, select Enable unattended installation, browse to the appropriate unattend file, and then click Open.
     
  4. Click OK to close the Properties page.
• Prestage a specific client with a client unattend file. You can use Windows Deployment Services to link physical computers to computer account objects in Active Directory Domain Services (AD DS). This is called prestaging the client. Prestaged clients are also called “known computers”. Prestaging a client allows you to configure properties on the computer account to control the installation for the client. For more information, see http://go.microsoft.com/fwlink/?LinkID=131987. Use the following procedure to prestage a client with an unattend file.
  
  
  To prestage a client computer with a client unattend file

  1. Click Start, right-click Command Prompt, and then click Run as administrator.
  2. To prestage the computer, use the following syntax: WDSUTIL /Add-Device /Device: /ID:.
     Example: WDSUTIL /Add-Device /Device:Computer1 /ID:ACEFA3E81F20694E953EB2DAA1E8B1B6
  3. To associate an unattend file with the prestaged client, use the following syntax where is the relative path to the unattend file from the Remote Install folder: WDSUTIL /Set-Device /Device: /WDSClientUnattend:.
     Example: WDSUTIL /Set-Device /Device:Computer1 /WDSClientUnattend:WDSClientUnattend\Unattend.xml
• Enable the Auto-Add policy and assign the client unattend file when you approve the installation. You can enable the Auto-Add policy to specify that unknown computers (computers that have not been prestaged in AD DS) require administrative approval before they will be allowed to boot against the server to install an operating system. Clients that are awaiting approval will be shown in the Pending Devices node of the MMC snap-in. If you approve the pending computer, the computer will continue booting from the network, and a computer account object will be created in AD DS to represent the physical computer. This is helpful because it gives you the ability to prestage computers without needing to know the client computer's GUID or MAC address in advance.
  
  
  To enable the Auto-Add policy and associate a client unattend file

  1. Right-click the server and click Properties.
  2. On the PXE Response tab, select For unknown clients, notify administrator and respond after approval.
  3. When an unknown computer boots to the server, the computer will appear in the Pending Devices node of the MMC snap-in. To approve the computer and associate an unattend file with it, use the following syntax where is the relative path to the unattend file from the RemoteInstall folder: WDSUTIL /Approve-AutoAddDevices /RequestID: /WDSClientUnattend:.
     Example: WDSUTIL /Approve-AutoAddDevices /RequestID: /WDSClientUnattend:WDSClientUnattend\Unattend.xml
     

     Note
     For more information about approving pending computers, see How to Manage Client Computers (http://go.microsoft.com/fwlink/?LinkID=115265).

Image unattend file

Next create an image unattend file, which automates the later phases of Setup (for example, offline servicing and Sysprep specialize). This file uses the Unattend.xml or Sysprep.inf format, depending on the version of the operating system of the image. See the Windows AIK for information about the settings in this file. For example files, see Sample Unattend Files (http://go.microsoft.com/fwlink/?LinkId=122642). After you have created the file, associate it with an image using the following procedure.

To associate an image unattend file

1. Create an Unattend.xml file (or a Sysprep.inf file for Windows Server 2003 and earlier versions of Windows) with settings applicable to the later phases of Setup.
   
2. The next steps depend on which file you create:
   
   • Save Sysprep.inf files. Save these files to the $OEM$ structure of the image (for example, D:\RemoteInstall\Images\Windows XP\winxpsp2\$OEM$\$1\sysprep\sysprep.inf). Now when you deploy the image, Setup will automatically locate and use the Sysprep.inf file.
     
   • Save Unattend.xml files. Save Unattend.xml files to any location, and then associate the file with an image. To do this, open the MMC snap-in, right-click the image that you want to associate with the unattend file, and then click Properties.
     
     On the General tab, click Allow image to install in unattend mode, click Select File, browse to select the unattend file, and then click OK twice. The Unattend.xml file will be saved to the following location: \RemoteInstall\Images\\\Unattend\ImageUnattend.xml.
     
     

After you have configured both unattend files, you are ready to deploy an image. When you initiate a network boot, the installation will proceed using the settings from both unattend files.

WDS: Creating Discover Images

Applies To: Windows Server 2008, Windows Server 2008 R2

A discover image is a type of boot image that you can use to install an operating system on a computer that is not capable of network booting using the Pre-Boot Execution Environment (PXE). You may want to use a discover image in the following scenarios:

• A client is not PXE-enabled.
  
• A client is on a different subnet and you do not have method of getting PXE to the client (for example, IP helper tables or Dynamic Host Control Protocol (DHCP)).
  
• You have many Windows Deployment Services servers and want to target a specific server.
  

When you boot a computer into a discover image, the Windows Deployment Services client will locate a Windows Deployment Services server, and then you can choose the install image you want to install. You can specify a destination server and the discover image will locate and install the image from the server automatically. You can also configure discover images to target a specific Windows Deployment Services server. This means that if you have multiple servers in your environment, you can create a discover image for each one, and then name them based on the name of the server.

Prerequisites for creating discover images

• A CD or DVD or flash drive to store the image.
  
• A Boot.wim file on the server. In most cases this should be the Boot.wim file from the product DVD, although you can also create a discover image using a custom boot image. For more information, see Creating Images (http://go.microsoft.com/fwlink/?LinkId=115311).
  
• A disk-burning utility if you are burning the image to CD or DVD.
  

Steps for creating discover images

The following procedures walk you through creating a discover image, and then creating media that contains the image.

To create a discover image

1. In the Windows Deployment Services MMC snap-in, expand the Boot images node.
   
2. Right-click the image that you want to use as a discover image.
   
3. Click Create Discover Boot Image.
   
4. Follow the instructions in the wizard, and when it is completed, click Finish.
   
5. Use the following procedure to create media that contains the image.
   

Note
You can also perform this procedure at a command prompt by running WDSUTIL /New-DiscoverImage /Image: /Architecture:{x86|x64|ia64} /DestinationImage /FilePath:. To specify which server the discover image connects to, append /WDSServer:.

To create media that contains the discover image

1. Download and install the Windows Automated Installation Kit.
   
2. Click Start, click All Programs, click Microsoft Windows AIK, and then click Deployment Tools Command Prompt.
   
3. To create a Windows Preinstallation Environment (Windows PE) build environment, type:
   CopyPE C:\Winpe
   
4. To copy the discover image that you created to the environment, type:
   Copy /y c:\discover.wim c:\Winpe\ISO\Sources
   
5. To change back to the PETools folder, type:
   Cd C:\Program Files\Windows AIK\Tools\PETools
   
6. To create the bootable .iso image, type:
   Oscdimg -n -bc:\winpe\ISO\boot\etfsboot.com c:\winpe\ISO c:\.iso
   
7. Use a utility that can create a CD or DVD to transfer the .iso image to the media.
   

   Note
   Using the Windows copy command to copy the image to media will not create a bootable disk.

   
   
8. Boot the client computer to the media. You may need to change the boot order in the client BIOS to do this. When the client boots to the media, your experience will be the same as if you had booted to the network.
   

WDS: Creating Custom Install Images

Applies To: Windows Server 2008, Windows Server 2008 R2

You can use Windows Deployment Services to create custom install images. To do this, use the instructions in this section to create a capture image, prepare a reference computer using Sysprep, and then capture the operating system using the Image Capture Wizard. These images provide an alternative to the command-line utility, ImageX.exe, which is included in the Windows Automated Installation Kit (Windows AIK). For more information including a comparison of capture images and ImageX.exe, see Creating Images (http://go.microsoft.com/fwlink/?LinkId=115311).

Note
Note that images that are older than Windows Vista are hardware abstraction layer (HAL)–specific, so you cannot deploy an image of one HAL type to a computer with a different HAL.

Prerequisites for creating custom install images

• Ensure that there is enough disk space to create and store the new images.
  
• Be a member of the Local Administrators group on the Windows Deployment Services server.
  
• Have a Boot.wim file on the server. In most cases this should be the Boot.wim file from the product DVD, although you can create a capture image using a custom boot image. For more information, see Creating Images (http://go.microsoft.com/fwlink/?LinkId=115311).
  

Steps for creating a capture image

To create an install image, you must first create a capture image. Capture images are boot images that you boot a client computer into to capture the operating system into a .wim file. You can also create media (a CD, DVD, USB drive, or other type of media) that contain a capture image, and then boot a computer from the media.

To create a capture image

1. In the Windows Deployment Services MMC snap-in, expand the Boot Images node.
   
2. Right-click the image to use as a capture image. In most cases, this should be the Boot.wim file from the Windows Server 2008 or Windows Server 2008 R2 product DVD. For instructions about creating custom boot images, see Creating Images (http://go.microsoft.com/fwlink/?LinkId=115311).
   
3. Click Create Capture Boot Image. Note that this will create a new image using the image you select (it will not overwrite your existing image).
   
4. Type a name, a description, and the location where you want to save a local copy of the file.
   
5. Continue to follow the instructions in the wizard, and when it is complete, click Finish.
   

   Note
   If you have Windows Server 2008 R2, you can select the check box on the last page of the wizard to add the image to the server (instead of following steps 6-8).

   
   
6. To add the image to the server, right-click the boot image folder and click Add Boot Image.
   
7. Browse and select the new capture image, and then click Next.
   
8. Follow the instructions in the wizard.
   
9. After you have created the image and added it to the server, follow the instructions in the next section to boot a computer into the capture image and capture the operating system.
   

Steps for creating an install image

Now that you have a capture image, you need to prepare a reference computer and then create the install image. The reference computer can be a computer with a standard Windows installation or a Windows installation that has been configured for your environment. First, you boot a computer (which has been prepared with Sysprep) into the capture image. Then a wizard creates an install image of the reference computer and saves it as a .wim file. After that, you can deploy the .wim file to a computer.

To create a custom install image

1. Create a reference computer (install the operating system, applications, and make any other changes that you want).
   
2. On computers running operating systems earlier than Windows Vista, ensure that you have the correct version of Sysprep.exe on the computer.
   
3. At a command prompt on the reference computer, change folders to \Windows\System32\Sysprep or the folder that contains Sysprep.exe and Setupcl.exe.
   
4. Run the command sysprep /oobe /generalize /reboot. If you prefer, you can also use the Sysprep graphical user interface by double-clicking Sysprep.exe.
   

   Note
   On computers running operating systems earlier than Windows Vista, run sysprep -mini –reseal -reboot.

   
   
5. When the computer restarts, perform a network boot on the computer by pressing F12.
   
6. In the boot menu, select the capture image that you created in the preceding procedure, and then click Next.
   
7. Choose the appropriate drive, and then provide a name and description for the image. Click Next to continue.
   

   Important
   You will see only drives that contain operating systems prepared with Sysprep. If you do not run the command in step 4, you will not see any drives to capture.

   
   
8. Click Browse and then browse to a local folder where you want to store the captured install image. We recommend that you use a local location to avoid image corruption in the event of a network malfunction, but you can specify a mapped network drive.
   
9. Type a name for the image, using the .wim file name extension, and then click Save.
   
10. Select Upload image to WDS server.
    
11. Type the name of the Windows Deployment Services server, and then click Connect.
    
12. If prompted for credentials, provide a user name and password for an account with sufficient permissions to connect to the Windows Deployment Services server.
    
13. In the Image Group list, select the image group in which you want to store the image.
    
14. Click Finish.
    
15. Now you can PXE boot a client computer to install this image.
    

Windows Deployment Services Getting Started Guide

Applies To: Windows Server 2008, Windows Server 2008 R2

This guide contains step-by-step guidance for how to install and configure the most common scenarios for Windows® Deployment Services (sometimes called “WDS”) in Windows Server® 2008 and Windows Server 2008 R2. Note that you cannot use Windows Deployment Services with the Server Core installation option. To download the Windows Deployment Services documentation (including a step-by-step guide, deployment content, and WDSUTIL command-line syntax), see http://go.microsoft.com/fwlink/?LinkId=89381.

Note
For information about Windows Deployment Services Update for Windows Server 2003, see http://go.microsoft.com/fwlink/?LinkId=66145.

What is Windows Deployment Services?

Windows Deployment Services is the updated and redesigned version of Remote Installation Services (RIS). Windows Deployment Services enables you to deploy Windows operating systems over the network, which means that you do not have to install each operating system directly from a CD or DVD.
For information about what is new or included in your version of Windows Deployment Services, see Windows Deployment Services: What’s New (http://go.microsoft.com/fwlink/?LinkId=140114).

Who should use this guide?

Windows Deployment Services is intended for deployment specialists who are responsible for the deployment of Windows operating systems. This guide assumes that you have a working knowledge of common desktop deployment technologies, as well as networking components such as Dynamic Host Configuration Protocol (DHCP), Domain Name System (DNS), and Active Directory® Domain Services (AD DS). The target audiences are:

• Deployment specialists interested in deploying Windows images to computers.
  
• IT planners, designers, or analysts evaluating Windows Server 2008 or Windows Server 2008 R2.
  

In this guide

• Quick start checklist
  
  
• Installing Windows Deployment Services
  
  
• Configuring Windows Deployment Services
  
  
• Installing an install image
  
  
• Upgrading from a server running RIS on Windows Server 2003
  
  
• Uninstalling or uninitializing Windows Deployment Services
  
  
• Advanced tasks
  
  
  • Performing Multicast Deployments
    
    
  • Managing and Deploying Driver Packages (Windows Server 2008 R2 only)
    
  • Creating Custom Install Images
    
    
  • Performing an Unattended Installation
    
    
  • Creating Discover Images
    
    
• Additional references
  
  

Quick start checklist

To install and configure Windows Deployment Services in order to install an operating system, perform the following steps.

 

	Task	Reference
	Install Windows Deployment Services. Note If you are upgrading your server, see Upgrading from a server running RIS on Windows Server 2003 .	Installing Windows Deployment Services
	Configure the server and add the default images (Install.wim and Boot.wim) that are included on the product DVD in the \Sources folder.	Configuring Windows Deployment Services
	Install an operating system.	Installing an install image

Installing Windows Deployment Services

Prerequisites for installing Windows Deployment Services

The following are requirements for installing this role, depending on whether you choose the default installation (both Deployment Server and Transport Server), or only the Transport Server role service.

 

Deployment Server and Transport Server

Transport Server

AD DS. A Windows Deployment Services server must be either a member of an AD DS domain or a domain controller for an AD DS domain. The AD DS domain and forest versions are irrelevant; all domain and forest configurations support Windows Deployment Services. DHCP. You must have a working DHCP server with an active scope on the network because Windows Deployment Services uses PXE, which relies on DHCP for IP addressing. DNS. You must have a working DNS server on the network before you can run Windows Deployment Services. NTFS volume. The server running Windows Deployment Services requires an NTFS file system volume for the image store. Credentials. To install the role, you must be a member of the Local Administrators group on the server. To initialize the server, you must be a member of the Domain Users group. For more information about this, see Required Permissions (http://go.microsoft.com/fwlink/?LinkId=115301).

For Windows Server 2008: The only prerequisite is that you must be a member of the Local Administrators group on the server to install Transport Server. Note also that a PXE provider is not installed with Transport Server, so you must create a custom PXE provider in order to network boot. For Windows Server 2008 R2: You must be a member of the Local Administrators group to install Transport Server. In addition, if you are using Transport Server to network boot, your environment must contain DHCP (Windows Server 2008 R2 contains a PXE provider, which allows you to network boot).

Steps for installing Windows Deployment Services

You can install Windows Deployment Services by using the Initial Configuration Wizard, Server Manager, or the command line.

• To install the role by using the Initial Configuration Wizard, click Add roles on the Initial Configuration Tasks startup screen. Click Next and then select Windows Deployment Services.
  
• To install the role by using Server Manager, click Add roles, which is located in the Roles Summary pane. Click Next and then select Windows Deployment Services.
  
• To install the role by using the command line, run one of the following two commands:
  
  
  • For Deployment Server, run ServerManagerCmd -install WDS.
    
  • For Transport Server, run ServerManagerCmd -install WDS-Transport.
    

During the installation, you have the following two role services to choose from. For a detailed comparison of these options, see Using Transport Server (http://go.microsoft.com/fwlink/?LinkID=115298).

• Deployment Server. To install this option, ensure that both Deployment Server and Transport Server are selected on the second screen of the installation wizard. This is the most common option because it provides the full functionality of Windows Deployment Services, which you can use to configure and remotely install Windows operating systems. Note that Deployment Server is dependent on the core parts of Transport Server.
  
  

  Note

  If you want to manage Windows Deployment Services on a remote server that is running Windows Server 2008 or Windows Server 2008 R2, you can install the Remote Server Administration Tools. To do this, open Server Manager, right-click the Features node, click Add Features, and locate Remote Server Administration Tools. This will install WDSUTIL and the Windows Deployment Services MMC snap-in on the server.

• Transport Server. To install this option, clear the Deployment Server check box on the second screen of the installation wizard. This option provides a subset of the functionality of Windows Deployment Services. It contains only the core networking parts. You can use Transport Server to create multicast namespaces that transmit data (including operating system images) from a standalone server. You should use this option if you want to transmit data by using multicasting, but you do not want to incorporate all of Windows Deployment Services. This guide focuses on the functionality of the complete installation of Windows Deployment Services (Deployment Server role service). If you choose to install the Transport Server role service, see Using Transport Server (http://go.microsoft.com/fwlink/?LinkID=115298).
  

Configuring Windows Deployment Services

After you install the server role, you must configure the server. Once you have used the instructions in this section to configure the server, add a boot image, and an install image, you will be ready to deploy images.

Known issues with configuring Windows Deployment Services

• Internet Protocol version 6 (IPv6) is not supported for Windows Deployment Services in Windows Server 2008. Windows Server 2008 R2 does not support IPv6 for network booting, but does support IPv6 for deploying images.
  
• If you are running Windows Deployment Services and a non-Microsoft DHCP server on the same computer, in addition to configuring the server to not listen on port 67, you will need to use your DHCP tools to add Option 60 to your DHCP scopes.
  
  

  Note

  There are some scenarios (particularly those that require running a DHCP server) that do not support adding custom DHCP option 60 on the same physical computer as the Windows Deployment Services server. In these circumstances, it is possible to configure the server to bind to UDP Port 67 in nonexclusive mode by passing the SO_REUSEADDR option. For more information, see Using SO_REUSEADDR and SO_EXCLUSIVEADDRUSE (http://go.microsoft.com/fwlink/?LinkId=82387).

• If DHCP is installed on a server that is located in a different subnet, you will need to do one of the following:
  
  
  • (Recommended) Configure your router to forward broadcast packets. All DHCP broadcasts by client computers on UDP port 67 should be forwarded directly to both the DHCP server and the Windows Deployment Services server. Also, all traffic on UDP port 4011 from the client computers to the Windows Deployment Services server should be routed appropriately (these requests direct traffic, not broadcasts, to the server).
    
  • Add DHCP options 66 and 67. Option 66 should be set to the Windows Deployment Services server, and option 67 should be set to boot\x86\wdsnbp.com. For more information, see Managing Network Boot Programs (http://go.microsoft.com/fwlink/?LinkId=115304).
    
• If Windows Deployment Services and DHCP are running on the same computer, configuring Windows Deployment Services to not respond to any client computers will not work. This is because although Windows Deployment Services will not respond, DHCP will. You can try to work around this issue by disabling DHCP option 60 on the DHCP tab
  

Steps for configuring Windows Deployment Services

To configure the server role, use the following procedure. Then see the following section to add images to the server.

To configure Windows Deployment Services

1. Ensure that you are a Domain Administrator.
   
2. Click Start, click Administrative Tools, and then click Windows Deployment Services. If there is not a server listed under the Servers node, right-click the Servers node and click Add Server to add the local computer.
   
3. In the left pane of the Windows Deployment Services MMC snap-in, expand the list of servers.
   
4. Right-click the server, and then click Configure Server (Note that the screenshots included in this document are from Windows Server 2008).
   
5. Follow the instructions in the wizard.
   
6. When the configuration is completed clear the Add images to Windows Deployment Services now check box and then click Finish.
   

Now that you have configured the server, you will need to add images. For instructions, see the next section.

Steps for adding images

You must add at least one boot image and one install image before you will be able to boot to the Windows Deployment Services server and install an image.

• Boot images. Boot images are Windows PE images that you boot a client computer into to perform an operating system installation. In most scenarios, you should use the Boot.wim file on the product DVD from one of the following operating systems:
  
  
  • Client: Windows Vista (with at least Service Pack 1 (SP1)) or Windows 7
    
  • Server: Windows Server 2008 or Windows Server 2008 R2
    
  You can also use custom boot images that you have created using the Windows AIK (for example, for diagnostic testing).
  
• Install images. Install images are the operating system images that you deploy to the client computer. You can use the Install.wim file from the product DVD to deploy images for Windows Vista, Windows Server 2008, Windows 7, and Windows Server 2008 R2. For operating systems released prior to Windows Vista, you must create a custom install image. For instructions, see Creating Custom Install Images and Deploying Earlier Versions of Windows.
  

To add the Install.wim from the product DVD, use the following procedures.

To add the default install image included on the product DVD

1. In the Windows Deployment Services MMC snap-in, right-click the Install Images node, and then click Add Install Image.
   
2. Specify a name for the image group, and then click Next.
   
3. Browse to select the default install image (Install.wim), which is located in the \Sources folder of the product DVD, and then click Open.
   
4. To add a subset of the images included in the Install.wim file, clear the check boxes for the images that you do not want to add to the server. You should add only the images for which you have licenses.
   
5. Follow the instructions in the wizard to add the images.
   
6. Click the image group to verify that the correct images were added.
   
7. Repeat this procedure to add any additional install images.
   

To add the default boot image included on the product DVD

1. In the left pane of the Windows Deployment Services MMC snap-in, right-click the Boot Images node, and then click Add Boot Image.
   
2. Browse to choose the default boot image (Boot.wim) on the product DVD, located in the \Sources folder.
   
3. Click Open and then click Next.
   
4. Follow the instructions in the wizard to add the image.
   
5. Repeat this procedure to add any additional boot images. When multiple boot images are available to client computers, clients will be presented with a boot menu that displays the boot images. For more information, see Managing the Boot Menu (http://go.microsoft.com/fwlink/?LinkId=115305).
   
6. If you want to modify any of the settings of the server, right-click the server in the MMC-snap in, and click Properties.
   
7. Now that you have at least one boot and install image on the server, you can perform a PXE boot on a client computer to install an operating system using the steps in the following section.
   

Installing an install image

After you have at least one boot and one install image on the server, you can deploy an install image.

Prerequisites for installing an install image

• The client computer must be capable of performing a PXE boot.
  
• Your user account must be a member of the Domain Users group.
  
• The client computer must have at least 512 MB of RAM, which is the minimum amount of RAM for using Windows PE.
  
• The client must meet the system requirements for the operating system of the install image.
  

Steps for installing an install image

To perform a PXE boot on a computer to install an image, use the following procedure.

To install an operating system

1. Configure the BIOS of the computer to enable PXE booting, and set the boot order so that it is booting from the network first.
   
2. Restart the computer, and when prompted, press F12 to start the network boot.
   
3. Select the appropriate boot image from the boot menu. (This boot image selection menu will be available only if you have two or more boot images on the server.)
   
4. Follow the instructions in the Windows Deployment Services user interface.
   
5. When the installation is completed, the computer will restart and Setup will continue.
   

Upgrading from a server running RIS on Windows Server 2003

This section applies to you if want to upgrade your server running Windows Server 2003 with SP1 or SP2. We recommend that you perform a clean installation whenever possible. However, if you decide that you want to upgrade your server, you should read the upgrade guidance in Upgrading to Windows Server 2008 (http://go.microsoft.com/fwlink/?LinkId=110832) to ensure that your upgrade is successful.
When moving an existing RIS infrastructure to Windows Deployment Services, we recommend that you upgrade all of your servers to Windows Deployment Services. The three server modes on Windows Server 2003 and the ability to convert RIPREP images enable you to transition seamlessly from RIS to Windows Deployment Services. We do not recommend installing Windows Deployment Services on new servers in your environment while continuing to maintain the existing RIS servers—because it requires additional hardware and administrative overhead. Having two PXE servers that are configured differently on the same network segment can lead to unpredictable results.
There are three modes of operation for Windows Deployment Services in Windows Server 2003: Legacy, Mixed, and Native. Your server must be in Native mode to upgrade to Windows Server 2008 or Windows Server 2008 R2. Your upgrade will be blocked if RIS is configured, or if your server is in Legacy or Mixed mode. To determine which operating mode the server is currently in, run the command WDSUTIL /get-server /show:config.
To determine how to upgrade, consider which of the following scenarios applies to you:

• If RIS is currently running on the server but you do not have Windows Deployment Services installed, you must install it before upgrading. Windows Deployment Services is included in the Windows AIK and Windows Server 2003 SP2. For more information about how to install and configure Windows Deployment Services, see the Step-by-Step Guide for Windows Deployment Services in Windows Server 2003 (http://go.microsoft.com/fwlink/?LinkId=94643). Then use the following procedures to switch from Legacy mode (the default setting) to Native mode.
  
• If RIS was installed on the server when you installed Windows Deployment Services, it will be in either Legacy or Mixed mode; you will need to change it to Native mode before upgrading.
  
• If RIS was not installed on the server when you installed Windows Deployment Services, the server will be in Native mode and therefore ready to be upgraded.
  

Use one or both of the following procedures to change the server mode to Native.

To change the server mode from Legacy to Mixed

1. Initialize the server by doing one of the following:
   
   • Using the MMC snap-in. On the Start menu, click Administrative Tools, and then click Windows Deployment Services. Right-click the server, and then select Initialize Server.
     
   • Using WDSUTIL. Run the command WDSUTIL /Initialize-Server /RemInst:C:\RemoteInstall (assuming that C:\RemoteInstall is the location of your REMINST shared folder).
     
2. When the process is completed, use the following procedure to change the server mode from Mixed to Native.
   

To change the server mode from Mixed to Native

1. Retire your RISETUP and RIPREP images, or convert them to .wim format. To retire them, just delete the images. If you want to convert them, you have two options:
   
   • Convert them offline (for RIPREP images only). To do this right-click the Legacy Images node, right-click the image and click Convert to WIM.
     
   • Deploy and recapture them by using the Image Capture Wizard (for RIPREP or RISETUP images). For instructions, see Creating Custom Install Images.
     
2. Run the command WDSUTIL /Set-Server /ForceNative.
   
3. When the process is completed, the server is ready to be upgraded. For more information about upgrading, see Upgrading to Windows Server 2008 (http://go.microsoft.com/fwlink/?LinkId=110832).
   

Uninstalling or uninitializing Windows Deployment Services

To uninstall the Windows Deployment Services component, you can click Remove Roles in Server Manager, or run ServerManagerCmd -remove WDS at an elevated Command Prompt window.
You can also uninitialize the server if you want to reset the server to a nonconfigured state but you do not want to uninstall Windows Deployment Services. This is helpful if you want to start over but would like to retain existing settings. For example, if you want to move the RemoteInstall folder (perhaps you got a new hard disk for your server and wanted to move this folder to it), you would uninitialize the server, copy the folder to the new location, and then reinitialize the server by using the new path. To uninitialize the server, run the command WDSUTIL /uninitialize-server in an elevated Command Prompt window.

Advanced tasks

For instructions on performing more advanced tasks, see the following topics:

• Performing Multicast Deployments. In order to deploy an image using multicasting instead of unicasting, you must first create a multicast transmission. Multicast transmissions make the image available for multicasting, which enables you to deploy an image to a large number of client computers without overburdening the network.
  
• Managing and Deploying Driver Packages. If you have Windows Server 2008 R2, you can deploy driver packages to client computers as part of an installation, and you can add driver packages to boot images prior to deployment.
  
• Creating Custom Install Images. You can use Windows Deployment Services to create custom install images using capture images, which provide an alternative to the command-line utility, ImageX.exe. To create a custom image, you create a capture image, prepare a reference computer using Sysprep, and then capture the operating system using the Image Capture Wizard.
  
• Performing an Unattended Installation. You can automate the entire deployment using two unattend files: one for the Windows Deployment Services user interface screens, and one for the later phases of Setup.
  
• Creating Discover Images. A discover image is a type of boot image that you can use to install an operating system on a computer that is not capable of network booting using the Pre-Boot Execution Environment (PXE).
  

Additional references

• For detailed deployment documentation for Windows Deployment Services, see http://go.microsoft.com/fwlink/?LinkId=81873.
  
• For a newsgroup about Windows Deployment Services, see Setup and Deployment at http://go.microsoft.com/fwlink/?LinkId=87628.
  
• To download a .doc version of this guide and other Windows Deployment Services documentation for Windows Server 2008, see http://go.microsoft.com/fwlink/?LinkId=89381.
  
• To see procedures for tasks that are not covered in this guide, see How to Perform Common Tasks (http://go.microsoft.com/fwlink/?LinkId=115268).
  
• To provide feedback about this documentation, e-mail wdsdoc@microsoft.com.