As new computers are purchased, before deploying them you install a standard software platform and enable the computer for future unattended reimaging.

1. Create a model workstation of each type that you'll deploy, and create an image of each on a ZfD imaging server. These images should include the Novell^® Client^TM and imaging agent.
2. Create imaging diskettes that point to the ZfD imaging server where the model images are stored.
3. Create a policy for unregistered workstations that specifies which image to put on a new machine, depending on its hardware.

As each new computer comes in, do the following:

1. Physically connect it to the network, boot it with the imaging diskettes and install the imaging partition.
2. Reboot from the imaging partition and let the computer be auto-imaged by the policy.
3. After deploying the machine, have the user register it as a Workstation object in NDS^®.

With minimal disruption to users, you enable existing workstations for possible future reimaging.

This might need to be phased in by local administrators. Each administrator could:

1. Upgrade each workstation to the latest Novell Client using Automatic Client Update.
2. Install the ZfD imaging agent on each workstation by distributing an Application object.
3. Register each workstation as a Workstation object in NDS.
4. Prepare a few sets of imaging boot diskettes or CDs that users can use when they run into trouble. These devices could point to an imaging server that contains the same, clean images used for brand new computers.
5. If a user runs into trouble, use the strategy for corrupted workstations below.

Without data loss or undue disruption to users, you fix workstations that have become misconfigured or corrupted.

1. Create a policy for registered workstations. Use the same image-selection logic as the policy for brand new (unregistered) workstations.
2. When a computer needs a fix, have the user back up (to the network) any files that he/she wants to keep.
3. Flag the Workstation object in NDS to receive an image the next time it boots.
4. Have the user reboot. If it's an older workstation (without a ZfD partition), the user should boot with the imaging diskettes or CD. If it's a newer workstation (with a ZfD partition), the user should boot from the hard disk.

After each lab session, you restore every workstation to a clean state, removing any changes or additions made during the session.

1. Create an image of a clean, model workstation and store it on a ZfD imaging server. The image should include the Novell client and imaging agent.
2. Create imaging diskettes that point to the ZfD imaging server where the clean image is stored.
3. Create a policy for unregistered workstations that specifies the clean image to restore. Choose the option to always force down the same base image.

Deploy each lab computer as follows:

1. Physically connect it to the lab network, boot it with the imaging diskettes and install the imaging partition.
2. Reboot from the imaging partition and let the computer be auto-imaged by the policy.
3. At the end of each lab session, reboot each computer and let it be auto-imaged by the policy.