Image Deployment

After your image is created, it is time to deploy it using a Deployment task sequence. Deployment task sequences are typically much more involved than Build and Capture task sequences because they incorporate all the various tasks to deploy a complete end-user system.

To create a Deployment task sequence, you must first have an image to work with that is imported into ConfigMgr. If you also plan to migrate user data, you must also have a USMT package built and available, as described in the "Creating the USMT Package" section. Creating an Image Deployment task sequence is nearly identical to creating a Build and Capture task sequence, as outlined in the "Automatic Image Creation" section; to create one, follow these steps:

1. Navigate to Software Library → Overview → Operating Systems → Task Sequence, and choose Create Task Sequence from the ribbon bar or right-click context menu to launch the Create Task Sequence Wizard.
2. Select Install an existing image package.
3. The remaining significant pages in the wizard define how to deploy a fully functional Windows system from a reference image.
   • Task Sequence Information: On this page, you define the Task sequence name and Description and set which boot image to use.
   • Install Windows: Choose the OS image to use, enter a product key (not required if deploying Windows 7 and using a KMS), and enable or disable the local administrator account. If enabled, you must also specify a password. By default, the task sequence is configured to Partition and format the target computer before installing the OS. If you are creating this task sequence for a refresh scenario and want to save user state locally, uncheck this check box.
   • Configure Network: Specify the workgroup or domain to join on this page along with the user account to use to join the domain.
   • Install Configuration Manager client: Choose the built-in ConfigMgr client agent installation package and add any additional installation properties. For deployments of workgroup systems, adding the SMSMP= property is highly recommended because the system is not joined to the domain and therefore cannot locate the MP or use AD.
   • State Migration: Options specified on this page configure the task sequence for a refresh scenario by including an initial set of tasks to capture user state and optionally capture Windows settings and network settings, and a final set of tasks to restore user state. You must specify the USMT package to use here as well as whether to Save user settings on a State Migration Point or Save user settings locally. If you selected to format and partition the computer system on the Install Windows page, saving the user state locally option is not available for selection.
     If you enable the capture of user state in the wizard, the resulting Capture User State task is configured to copy user state locally. As discussed in the "User State Without SMP" section, hard-linking is generally preferred, so you may want to update this task after the task sequence is configured so hard-linking is used instead.
     If you plan to use this task sequence for a replace scenario on destination systems, the condition placed on the Capture Files and Settings group will prevent it from running because the task sequence is initiated in WinPE and not from a current OS. For a task sequence that captures only user state on the source system in a replace scenario, delete the Install Operating System and Setup Operating System groups.
   • Include Updates: Choose whether to include updates in the reference system image. For updates to be considered, they must be in a deployment that targets the same collection also targeted by this task sequence's deployment.
   • Install Applications: Select applications to add to the reference image.

NOTE:
There are two basic types of boot images available: one for each major Windows architecture type. Which should you use? To deploy a Windows image using a Deployment task sequence, you can actually use either regardless of the architecture of Windows you are deploying. For a Build and Capture task sequence that directly references OS source files in the form of an OS Installer, the boot image used must match the architecture of the OS source files.
The primary point here is that using the x86 boot image for both x86 and x64 OS deployments is generally practiced by most organizations because it means not having to maintain both flavors of the boot image and eliminates any confusion over which boot image is being used or delivered to the client using PXE. You must still use the x64 boot image in x64 Build and Capture task sequences, but this limited use of the boot image makes it easier to manage.
The confusion arises when using unknown computer support in combination with available (nonrequired) task sequence deployments. In this case, ConfigMgr does not know which boot image to deliver to the client and chooses the boot image associated with the task sequence deployment with the highest assignment ID. This typically corresponds to the task sequence most recently deployed. This behavior is technically undefined, meaning it is not documented making it subject to change without notice.