Using PC Cloning Technology To Speed Networked PC Deployments.

Once, the deployment and configuration of new PCs was one of the most time consuming tasks of a reseller or systems integrator. However, a new method with some new technology now makes a different approach to disk cloning possible and the time it takes to configure a new PC can be reduced from hours to minutes.

Disk cloning lets you take a snapshot image of a source computer's hard drive and transfer that image to another computer. This snapshot is not just a copy of files and directories, but a detailed copy or image of everything on the drive, including OS file formatting, FAT tables, drive partition information, and so on. This image of the drive can be transferred to another computer, making its drive contents an exact replica of the original computer. In the following sections, we'll quickly describe how and why the process works. (The Fig shows each of the major components of a network disk cloning system.)

Imaging Software

The imaging application is a program that creates an image file and uploads it to the network server. Put simply, an image file is a file that contains a "picture" or "image" of one or more hard-drive partition(s) of a "source" PC. Any fully installed and configured PC may be used as a source machine from which to create an image file. If the source PC has been configured with applications and data files in addition to the operating system, then the image file will include both the operating system and its associated applications and data files.

When uploading an image file from the source machine onto the server, the imaging software first prompts the user for a filename and target folder (directory) where the image file will be stored. Typically this will be the folder on the network server or shared device where the master library of images is stored. The imaging software then displays the partition table of the source PC, allowing the user to select the desired partition(s) to be included in the image.

The imaging software is also invoked during the download process. When downloading an image file, the imaging software searches the database for the Alias/Name of the PC being configured. If the specified Alias/Name is not found, the agent automatically brings up a menu that allows the user to assign or randomly select an Alias/Name.

Once the Alias/Name is selected, the imaging software determines which image is to be downloaded and whether or not to image the target machine as a Sender or Receiver. The first PC to come up as "live" during an image broadcast session will be assigned as a Sender and all subsequent PCs will become Receivers. Senders receive the image from the server, write the image to the designated partitions, and resend the image to all Receivers. Receivers simply write the image to the designated partitions. The imaging software's final function is to reboot the now cloned PC.

Client Configuration Agent

The client configuration agent can run manually or it can be included in the image file, installed as part of the cloning process by the Imaging Software to run as a service following the boot sequence. This client configuration agent requires the use of a source file, which is placed on the hard drive by the imaging software during the imaging process. The source file contains specific configuration information such as IP address and computer name, which is unique to that PC. If this source file is found, the configuration agent makes changes to the registry as specified by the source file. If the source file is not found, the client configuration agent prompts the user for the required information.

When its task is completed, providing no errors have been logged in the error log file, the client configuration agent deletes the source file, its own executable, and the error log file. Optionally, the client configuration agent updates the status for the specific PC being configured in the Name/Alias object in the status database.

The client configuration agent displays critical error and informational messages on the screen while it is running, regardless of the mode. Other errors and messages are written in a .LOG file at the root of the partition containing the system root directory.

Configuration Database/Status Database

The configuration database is a set of data files on the server that contain specific configuration information for each PC to be configured. It stores configuration values such as IP Address, Computer Name, and other configuration values, which are unique for each target machine. The configuration database also functions as a status database providing reportable status information to the disk cloning console on each individual PC while the PCs are being cloned and configured. The configuration database is created and managed using the

disk cloning management console.

Management Console

The management console provides the interface where the vast majority of the user interaction will be performed. Its initial function is to allow the user to create a configuration SET. A configuration SET defines a group of machines to be configured and includes Names/Aliases and configuration values for each PC in the configuration set. Each Name/Alias within the "Set" contains the IP Addresses, Computer Names, Domain Names, Product Names, Product Numbers, and related information that may be required by the Client configuration agent when actually performing the client configuration task for the target PC.

The Management Console displays the current status of each Name/Alias in a given SET while the download and configuration processes are running. This allows the user to quickly view the progress of each PC as it is being imaged and configured.

Partition/Disk Image Files

The Imaging Software creates the image files and contains a snapshot or picture of one or more partitions of the source machine's disk drive. Typically, these files are compressed and stored on a network file server and become part of a library of image files, which the imaging software can use to install and configure a target PC at any time. Image Files also provide a convenient way to re-image a PC in the event of a major failure as part of a disaster recovery strategy within an IS department. The following sequence of operations outlines the steps an integrator might perform in order to use disk cloning technology to fully deploy systems within an organization.

Step One: Create An Image File

A source PC should be configured to contain the desired Operating System, Applications, and any associated Data File, if required. The Client Configuration Agent should also be installed on the source PC along with the Security ID generation (SIDgen) tools for NT if the image will be used to install and configure NT machines.

This PC is then rebooted, using a bootable DOS floppy. A network connection is established with a drive letter mapped to the folder (directory) on the server where the image file will be stored. The Imaging Software is then used to create an image file of the source PC's hard drive partition(s).

Step Two: Create The Configuration Database

Once an image file has been created and saved on the network server, the manager uses the disk cloning management console to create the configuration database. As described above, the configuration database contains all of the unique configuration information for each of the target PCs as part of the configuration SET. During this step, a connection is made between the configuration database SET and the image file created during step one. It should be pointed out that step one and step two can be accomplished in any order.

Step Three: Create The Client Boot Diskettes

As stated previously, the client configuration agent runs as a DOS process on each of the target PCs. The third step of the deployment process requires that DOS boot diskettes be created for the target PCs. These boot diskettes will contain the network drivers, protocol stacks, and network control files necessary to boot the machine, connect it to the network, and map a network drive to the location where the image file is located.

Step Four: Download The Desired Image

The next step in the sequence consists of booting the target PCs with the DOS boot disks created in step three, downloading the image file, and writing it to the target PC's hard drive. As stated previously, this process is actually accomplished by the imaging software running as a DOS process on each of the target machines.

During this part of the operation, the boot diskette will cause each of the target PCs to attach to the network server, map a drive letter to the location where the image file is located, and map a drive letter to the location where the cloning programs are located. The imaging software is then executed. This imaging software then checks the configuration database (created in step two) to determine what image file should be used and what the configuration values are that should be assigned to the target PC. It then copies the assigned image to the target PC and writes the configuration values to the target drive for later use by the configuration agent. As progress continues, status is written back to the status database and reported to the management console.

Step Five: Boot Client PCs To Run The Configuration Agent

Prior to the final step of the imaging/configuration process floppy disks must be removed from each of the target PCs. This allows each target PC to boot to the hard drive after the download process described in step four is complete.

Providing the client configuration agent was included in the image file, it will run automatically following the boot sequence. The PC's registry file will be automatically updated with the configuration values assigned to that machine from the configuration database. Some changes to the registry require that the target machine be rebooted in order for the changes to become active. Should this be the case, the client configuration agent will cause the machine to perform one final reboot in order to complete the configuration process.

If access to the network is still possible following final configuration of the target PC, which will be the case in most environments, the client configuration agent performs one final update to the status database in order to report that the target PC has been fully configured. Should any errors be detected during this final step, they are reported to the status database and written to an error log on the target PC's hard drive.

In the case of PCs running Windows NT, it's important to create a unique security ID (SID) for each computer as part of the final configuration process. This is a critical step in the imaging process. Otherwise, all NT machines would have identical SIDs, thus compromising standard Windows NT security.

One job remains in this final configuration step: that of having the client configuration agent remove itself, along with the configuration values from the target PC's hard drive. This avoids problems that could result from inadvertently restarting the client configuration agent at some future date.

Two of the main reasons IS professionals in corporations, configuration centers, and value added resellers use cloning software to deploy computers are to maintain configuration consistency and to save time on individual computer setup.

Even though cloning software handles general configuration issues as they relate to operating systems and user applications, as mentioned in the final step of the cloning process, it's critical to use a solution that also addresses the need to configure a computer's unique properties. For example, Windows 95-based computers rely on unique settings such as Computer Name, Workgroup Name, IP Address, IP Gateway, IP subnet mask, etc. Configuring each computer's unique settings can be very time consuming. This problem is exacerbated when configuring PCs with NT. Since cloned PCs are an exact duplicate of the original source PC, all cloned NT PCs will have the same SID.

At first glance, configuring the computers' unique settings may not seem like a big deal. After all, it may take only about five minutes to configure one computer's unique properties. However, five minutes per computer becomes very significant when deploying large quantities of computers on a regular basis. For example, twenty-five computers per week can add multiple configuration hours. Systems integrators that do over a hundred computers for that same time period can waste more than eight-and-a-half hours per week. For this reason, it's critical to choose a cloning solution that gives IS professionals the means to automatically configure each target computer's unique settings without physically going to each individual computer.

The introduction of computer cloning solutions makes a major impact on IS productivity. It has revolutionized the way new computers are deployed. With effective disk cloning technology, not only does adherence to PC deployment standards become much easier, but it also becomes possible to deploy new computers in a fraction of the time that was previously required. Used effectively, cloning can significantly reduce the Cost-Of-Ownership and increase the Return-On-Investment for resellers and integrators doing system deployments on any scale.

Jan Newman is the president of Altiris, Inc. (Lindon, UT).