Desktop layering is a method of dividing a disk image into logical parts that can be managed individually. This method can reduce the amount of storage necessary to back up a large number of disk images (See: image-based backup) and allow easier patching and disaster recovery. Layering can be applied to local physical disk images, client-based virtual machines, or host-based desktops.
The Windows operating system was not designed with layering in mind. As a result, every layering vendor must engineer its own solution. Some solutions manage an individual component of an image as its own layer, while others manage the entire disk image as a collection of layers. Typical divisions of a disk image can include user personalization layers, user installed application layers, departmental application layers, the core operating system, and driver layers to interact with hardware.
In traditional environments, disk images must be stored and managed on an individual basis. With desktop layering, if every user is running the same operating system, then the core operating system only needs to be backed up once for the entire environment. In addition, patches and upgrades can be made just once and pushed out to all the users, instead of patching all users’ machines individually. Company-wide applications can be included on the core operating system layer as well. Only the layers that contain user data and files need to be managed and backed-up on an individual basis. Other layers, such as departmental application layers or driver layers, can be managed on a per-department or per-hardware configuration basis.
When a new disk image is created, components can be assembled by choosing from different available versions of each layer. If a disk image becomes corrupted, administrators can replace or roll back only the affected layer, without disturbing the rest of the image.