Workstation Imaging System

The ZENworks for Desktops 4 system for imaging workstation is made up of the following components: Linux operating system, Linux imaging software, Imaging server agents, and the Windows workstation image-safe data agent. In addition to these components, objects in eDirectory, and some administrative tools in ConsoleOne, are there to get the job done.

There might be a Linux partition placed on a workstation (minimum of 15MB) that can hold the Linux OS and the imaging engine. It's ideal to place this partition on workstations that you might want to image on the fly (from ConsoleOne) because, when requested, the workstation gets notified of this work and then performs the imaging task. If no work exists for the imaging system to perform, the partition boots into the normal operating system. See the section on placing a Linux partition on a workstation for potential issues and instructions.

The ZENworks for Desktops 4 imaging system is designed to function in an automatic mode (although it does have a manual mode). The expectation is that you use the system to deploy images to your workstations in the network in order to set up initial systems or repair systems and get them back online as quickly as possible. Consequently, the system assumes that a workstation that is contacting the imaging server, but is not registered in the tree, is requesting an image; and a registered workstation is contacting the imaging server to determine whether any work (indicated by flags in the workstation object) is to be done.

The ZENworks for Desktops 4 product ships with the ZENworks Preboot Services product. This allows you to have PXE (Pre-Execution Environment) support right out of the box! Now you don't have to have a Linux partition on the workstation. Instead, you can boot PXE. The PXE Server provided in ZfD4 will then send the device the Linux Kernel. The imaging system will contact the imaging server for work and proceed to send and apply the image.

Linux Operating System

ZENworks for Desktops 4 uses the free Linux operating system to take and apply images to workstations. The imaging process begins by booting the workstation with some boot floppies that contain the Linux operating system and several drivers to handle most network cards. If the boot disks cannot support your network card, you need to look for those resources on the Internet. The following sites might have the driver you're looking for:

cesdis1.gsfc.nasa.gov/linux/drivers/index.html

www.linuxvoodoo.com/drivers/cards/nics/index.html

www.scyld.com/network/index.html

Linux Imaging Software

This is the Linux software application that actually takes and applies the image. This software is automatically started with the boot disks and is launched when the workstation is told to take or receive an image.

This software is not writing bits and bytes on the sectors, but it has knowledge of the various supported file system types, and reads and writes the files. The supported file system types are FAT16, FAT32, NTFS 4 (Windows NT), and NTFS 5 (Windows 2000/XP). Because the imaging software is reading files, it also writes files. This means that if you take the manual approach to restoring the image and the partition is not empty then you have a mingling of currently existing files with the files from the restore causing a very unexpected behavior (OS and driver files are intermixed, and so on).

Something else to note is that the images always have the suffix of .ZMG and are not compressed. You must have enough room on the destination server to store the entire image or the image transfer will fail and the partial image will be deleted.

The imaging software can function in one of two modes: automatic and manual. In the automatic mode the imaging software contacts the imaging server agent and requests any work. This work can be to take or receive an image (this is set in the workstation object). If an image must be taken or received, the imaging software begins the process. If no work must be done, the imaging software completes and the workstation reboots to the native operating system.

If the imaging software is in manual mode, the software does not automatically communicate with the agent, but places prompts on the workstation screen that enable you to perform specific partitioning and imaging tasks including taking or receiving an image. See the "Advanced Imaging" section in this to learn about the commands you can perform in the manual mode. When the imaging software is done, the user must manually request the reboot to the native operating system.

Imaging Server Agents

This is the agent that runs on the server and is responsible for communicating with the imaging software that is running on the workstation. These agents tell the workstation whether to take the image or to receive an image and are responsible for walking the eDirectory tree to find the image. This agent is affected by the Imaging Server Policy in the Server Policy Package. It is responsible for receiving information from the workstation and processing the rules in the policy in order to find an image object that should be applied to the workstation.

Once the image is determined, this agent gets the image file and transmits that image to the Linux image software residing on the workstation. It is also responsible for receiving any images that the workstation is sending it and storing them on the server in the specified and approved locations.

The image server is loaded on the NetWare server as imgserv.nlm and the NT/2000 version is a service DLL with the same name. The imaging server has a status screen that tells you some information on the number of requests and images it has received and served. It, unfortunately, does not have any information on the screen on currently receiving or delivering work. You can load the service with a debug option and get it to write a log file called ZIMGLOG.XML on the server.

Windows Workstation Image-Safe Data Agent

This agent resides on the Windows workstation and is responsible for receiving image-safe data from the disk and placing that into the Windows Registry. It also makes sure that the information in the Windows Registry is synchronized on the disk.

A special sector on the disk is reserved for placing information that is preserved despite having an image applied to the workstation. This way, a workstation keeps its IP address (or DHCP settings), computer name, workstation object, domain, and group names.

eDirectory Objects

Several objects are introduced to the tree in order to support ZENworks for Desktops 4 imaging. These objects are the following:

• Workstation Imaging Policy This policy is in the Workstation Policy Package and determines, for the associated workstations, the image to be used when a re-image is requested. See Chapter 9, "Setting Up a Workstation Policy Package," for more information about this policy.

• Imaging Server Policy This policy is in the Server Policy Package and determines the image used when an image is requested for a non-registered workstation. See Chapter 12, "Creating a Server Policy Package," for more information about this policy.

• Image Object This is an eDirectory object that represents an image taken of a workstation and stored on the imaging server. See the following section to learn more about the Image object.

• Workstation Object: Imaging Configuration These are some configuration parameters that are part of the workstation object. In this page, you can configure whether the workstation should take or receive an image on the next reboot.

Image Object

The administrator creates the Image object in the directory and associates it with an image file (.ZMG) taken by the ZENworks for Desktops 4 imaging system. To properly configure an Image object, you must do the following:

1. Take an image of a workstation and store that image on an imaging server. See the section that follows about creating a workstation image.

2. Launch ConsoleOne and browse to the container where you want the Image object to reside.

3. Create an Image object by selecting the container and then choosing the File, New, Object menu and selecting Workstation Image object from the list of objects.

4. Once the object is created, go into the object's properties by either selecting additional properties in the create process or by right-clicking the object and choosing Properties.

5. Select the Image Files tab and specify whether you want to use a base file image or set up scripted imaging. If you select standard imaging, you must administer the location of the .ZMG files by pressing the browse button in the Base Image File field. You then need to browse to the imaging server and then to the file system on that server to specify the .ZMG file. If you select scripted imaging, you can specify an imaging command in the text window that appears (see "Advanced Imaging" later in this chapter).

6. Choose the file set to bring as part of the image by selecting the set in the Use File Set parameter at the bottom-left side of the screen. You can have up to 10 file sets. See following section, "Modifying an Image," for more details.

7. If you are using standard imaging, you can also append additional images to this image by pressing the Add button and placing them in the Add on Image File field. These additional images are included with the base image and placed on the workstation when it is imaged. These images can be Application Object images (see Chapter 6, "Creating and Using Application Objects," for more information).

8. Press OK to save the administrative changes and get out of ConsoleOne.

Administrative Tools

Several tools exist that are available to perform the imaging operation. First are the snap-ins into ConsoleOne that enable you to create image objects and launch tools to create the imaging boot disks and to view and manipulate an image file.

The boot disk creator and the Image Explorer program are both Windows programs that are launched from the Tools menu of ConsoleOne. The boot disk creator creates the Linux disks needed to boot the workstations to communicate with the imaging server and to take and put the images.

The Image Explorer enables you to view the contents of the image file and to mark files in the image to be included in various sets of files, as well as remove and add files from/to the image. See the "Modifying an Image" section later in the chapter for more details.

Setting Up Workstations for Imaging

The ZENworks for Desktops 4 imaging engine that performs the actual workstation imaging is a Linux application. Therefore, the workstation must be booted to Linux temporarily while the imaging is performed. For a workstation to use the ZENworks engine, you need to prepare a bootable device that has the following components:

• Linux Kernel Either a bootable device or partition with Linux installed.

• Imaging Engine ZENworks for Desktops Imaging Software.

• Network Drivers Drivers to access the network to communicate with the ZENworks for Desktops imaging server.

You can use any of the following methods to prepare workstations to be imaged:

• Preboot services (PXE)

• Boot disks

• Boot CD

• Linux partition on a workstation