I was having an IM conversation with Brian at CMS Consulting about some deployment work and how IT departments support the desktops computers in their environment. A few days later Jason Cotton, a senior consultant at CMS Consulting Inc., sent the following in. It is a great article with useful information to help you improve your desktop environment.
As a consultant, I've had the opportunity to see how many organizations support their desktop environments. How systems are supported mainly depends on what investments a company has made in its desktop management tools and technology. While companies may have different styles of managing the desktop, the goals are usually the same; Minimize costs and keep users running. It is not always apparent how technology improvements and methodologies can significantly improve desktop management. Enhancing the productivity of the desktop environment can be accomplished simply by introducing more modern and automation-based tool sets.
Let's take a look at some options that can help improve your desktop environment.
Having dozens of images in use within an organization can cause any number of issues. The deployed image a customer receives might work, but support and troubleshooting times can increase drastically unless the image is consistent, documented and repeatable. Since this is absorbed in help desk costs, the consequences are often hidden.
If you know me, you know I've emphasized the Microsoft Core Infrastructure Optimization model and how it applies for desktops. The message is Basic, Standardized, Rationalized, and Dynamic, where each of these would represent the maturity of the desktop deployment model. The closer to the dynamic, the more effectively an organization can adapt to changing business needs. Getting there involves implementing the right technology solution, as well as improving and following the processes that have been developed through the project.
Projects completely redesigning all aspects of a desktop management are rare, but they do happen. Otherwise, a series of smaller projects over a longer period of time are also capable of delivering the same results, albeit while managing two modes of operation. As one area is improved, a neglected area may then show a gap that would become the next target for improvement. For instance, you complete an image consolidation project but do not have a system to deploy applications. Obviously, application management is a target where you could now save more money and deliver applications more effectively to your customers.
So, planning ahead and having goals that are part of a larger vision is very important. In this way, a basic and unmanaged desktop environment could be transitioned to dynamic without redesigning previously implemented projects.
For instance, if you just consolidated your images, but all non-base image applications are installed manually, it would make sense to implement a solution that would automate this task. In this example, applications could be supported and managed separately without affecting other aspects of the desktop management. Note that this solution did not introduce a managed process for introducing new applications and one would need to be created. Improving upon or introducing processes is good in that it delivers a method to standardize how packaging, deployment, or updates would be made.
The effort required to consolidate images down to a single image varies depending on the state you are in ? Basic, Standardized, Rationalized, and Dynamic. If some of these techniques are already being used then you may have momentum to continue.
You may be able to consolidate if the following scenarios are applicable.
- There is a delta due to organizational structure
- Multiple groups performing the same functions
- Segmented desktop management, thus a variety of versions present in each part of the organization
- Technology tools for desktop management differ
- Manual installs
- Automated installs
- Inconsistent application packaging
- Lines of business determine which applications are present
- Hardware types affect what software is loaded
- Hardware/Device drivers affect number of image permutations
- Policies vary for some organizational reason between images
- Different operating system versions or service pack levels
These bullets identify some typical causes for concern in managing multiple desktop images. Each situation can cause a new OS image to be built; combine this with the same situations in this list, combined with maintenance and updates, and soon an image environment is not really managed at all. It becomes "just make it work". Any post-deployment issues would be masked as help desk incidents and are not seen as something that could be prevented.
Looking at the consolidation options, we can break these down and look at why these areas are not just consolidated as a matter of fact.
Organization management style can only be improved when an organization recognizes or decides to structure desktop management accordingly. This topic can come up when an IT department looks towards upgrading to the next operating system.
Where desktop management tools are not the same, the image format can differ. Two business areas may follow the same process yet use different technologies to implement the process. Standardizing on one tool can be beneficial for the same reasons as having all of your desktops on the same operating system and service pack level; it would likely save support and licensing costs.
Keeping processes the same across an organization can reinforce the technology. The recommendation I would have is to have your technology assist in ensuring processes are followed correctly. Having organizations agree to use the same processes make the environment easier to support in the long term.
Differences among lines of business can be handled quite easily using modern desktop management tools. Combine this with a logical grouping of devices that represents your business, and you have a piece of technology that can be used in the process of defining which of those lines of business get installed during deployment. Figuring out how many lines of business you have, and what the application and configuration differences are, is possibly the most work in a consolidation project, because it requires detailed conversations with the business areas.
Historically HAL dependency was another issue, a typical example is laptops using one HAL type and desktops using another. As a result, organizations had two images for these HAL differences. Windows 7 no longer has this dependency, and therefore one image can be applied to a machine irrespective of its HAL type assuming the architecture is the same (x86 or x64).
Device drivers driving image updates is common since the hardware lifecycle today is about 9 months, and the operating system lifecycle is many times that. The fact is, there will be maintenance requirements to support new hardware. Today, there are methods that can load device drivers when the operating system is deployed, yet they do not have to be built into the image. By doing this, the core image does not need to be updated each time hardware changes. Instead, they can be stored separately and managed independently of the core image. This simplifies testing and improves QA since only the device driver library needs to be updated.
The technology policies varying across an organization is also common, however applying policies in the infrastructure and not directly in the image build is the recommended approach. Active Directory would be the best place to start by not building group policies into the image, the image is simpler and requires less maintenance. The group policy would be applied to all devices as soon as they join the network and is an example of a modular approach to building an image. Keeping components out of your base image and managing them outside the thick image is a great way to automate an image build. Of course building a modular or thin image like this comes at the cost of install time. Therefore, choosing the thick, thin, or a hybrid approach is important to minimize the need to rebuild a base image.