In a previous article, I explained how to create DeployStudio-compatible NetBoot images. You can use these images to cold-boot client computers into a network-based disk image, which in turn connects to the DeployStudio server to perform any number of deployment operations — from OS to software and settings or push out an entire thick image.
This time around, I’m going to cover the process of capturing an image from a pre-configured workstation — specifically, an Apple computer that has OS X, all updates, important software installed, and is configured properly for the environment. This workstation will serve as the reference or master that’s used to create a compressed master image that will be stored in the repository on the network server and may be used for later deployment to any number of Mac computers.
Before diving into the image capture process, let’s first review the requirements:
- Apple computer running OS X Server (10.6.8+)
- DeployStudio installed and server service configured
- NetInstall service configured
- NetBoot created using DeployStudio Assistant
- Switched network
- Static IP
- Broadband internet access (Ethernet)
- Apple computer with desired OS, updates, and applications installed (Master)*
* Note: Thick images are referred to as images that are created with all the relevant updates, software applications, and settings configured. These images, while typically being larger in size, are useful in that they contain everything that’s required by the organization in order for employees to be productive. However, best practices indicate that antivirus software, management end-points, and binding to directory services should not be included, as these tend to run afoul on the images.
Follow these steps to capture an image with DeployStudio.
- Launch DeployStudio Admin.app and authenticate using administrative credentials (Figure A).
- Clicking on Workflows will bring up the saved workflows, which are a grouping of tasks to be executed by the DeployStudio service. By default, certain workflows are preconfigured, such as Create a master from a volume, which will be used in its default format to create a deployable master image from the master computer (Figure B).
- On the master computer, power on the node and hold down the [option] key to enter the boot menu. Once loaded, a choice of available boot disks will appear, including network-based disks. If the network-based disks are not present, check OS X Server to verify that the NetInstall service is on and that the previously created NetBoot image is set as the default image. Additionally, if it’s the first time launching the boot menu, press and hold the [N] key, as that will prompt the computer to search the network for available servers running NetInstall and present you with available NetBoot disks. Once you’ve selected the correct NetBoot image created by DeployStudio, the computer will begin to boot over the network. This process may take a few minutes, depending on the speed of the switched network (Figure C).
- Once the NetBoot image is loaded, the DeployStudio Runtime will load and login automatically to the DeployStudio server, if that was an option selected during the NetBoot image creation process. If not, you may be prompted for admin-level credentials once again to authenticate to the server.
- After authenticating, a list of workflows will appear. When new workflows are created, they’ll also appear in this menu for easy selection. For now, manually select the Create a master from a volume, and click the triangle indicating play (Figure D).
- On the next window, you’ll be prompted to provide an image name (required) and keywords (optional). While anything may be entered as a name, I recommend a combination of the OS version, type of image being created, and a creation date. Try to keep the image name concise but recognizable at a glance. Keywords could be used as needed to further break down information contained, such as applications included or configurations made for specific offices or users. After this information has been entered, press the triangle button again to begin the image creation process (Figure E).
- Imaging or cloning takes take. Sure, it would take more time to manually configure each computer separately, but the initial process of image creation — when done properly — does need to run its course, especially when thick images are concerned. The more data being captured, the longer the process will take to complete. During this time, it’s best to just move on to another task, as the process will return back to the workflow selection screen once it’s complete (Figure F).
- Upon successful completion, navigate to the root of the DeployStudio repository that was created when DeployStudio was installed, and you’ll find a folder titled Masters. Another directory inside includes HFS, which is where all OS X-based images are stored. Inside that directory, you should find a .DMG file with the image name assigned to it in step #6. If the image size does not match up to the original computer’s HDD, don’t fret — DeployStudio tries to use compression whenever possible, since images can be rather large in size. Additionally, if imaging 10.7+, there may be a separate .DMG file for the recovery partition. DeployStudio creates a backup of the recovery partition by default (Figure G).
- With the image successfully created, the master station may now be rebooted or powered off. In the event that the master image boots back into NetBoot by default, simply click on the Apple menu and select Startup Disk from the list. The familiar startup disk launcher will load, where you can select the proper startup disk and then click Restart to reboot the machine into the correct disk (Figure H).
With that, the reference image is captured and stored in the server’s repository for future deployments. While the task is complete, I strongly urge testing the newly captured image on another Apple computer. Testing not only provides verification that the image captured is bit-for-bit accurate, but in the event that certain issues present themselves, it may provide validation of these errors and the opportunity to repair/correct them prior to deploying the image on a mass-scale.
Once you start utilizing images for deployment, it simply doesn’t end there. As time passes, the images themselves become outdated due to changes in business practices, software updates, and OS upgrades. However, having a good, stable image always allows a system admin to work on a solid foundation, modifying only the relevant pieces as needed, similar to delta updates on mobile devices.