IT Management Reference Guide

This section was written by Jeff Rohloff.

This article describes my experiences with researching, proposing and initiating efforts to develop a configuration management process for the Las Vegas Metropolitan Police Department (LVMPD). Configuration management is described and lectured on by academics and IT consultants alike, with different concepts and approaches on how an IT organization can implement such a process. My intent with this first part of the series is to show how we arrived at a common definition of the term configuration management and to describe the approach I used to propose a new method of performing configuration management. At the conclusion of part one I will share some important lessons learned from these activities.

In part two of this series I will describe how the LVMPD-ITB initiated efforts to begin designing a configuration process, how the Information Technologies Bureau (ITB) of the LVMPD was currently gathering and managing configuration data, and some of the lessons we learned from these efforts.

Definitions of Configuration Management

I began my efforts of implementing a configuration management process by assembling a cross-functional team. One of our first tasks was to agree on our definition of configuration management. This term means different things to different organizations. Research can introduce one to theories, and a myriad of concepts on how this process should be structured within an IT organization. However, I believe that in order to best understand the surreptitious nature of this process, it is necessary to share some of the varieties of definitions of configuration management. I offered the team three different, but well-accepted, definitions of configuration management from three very different sources.

The first was a definition defined by Reference Guide host, Rich Schiesser, who writes that configuration management is a process to ensure that the interrelationships of varying versions of infrastructure hardware and software are documented accurately and efficiently.

Dennis Drogseth, Vice President of Enterprise Management Associates offered a second version by describing configuration management as a process to identify and arrange network resources to optimize the quality of the user experience.

The IT Infrastructure Library (ITIL) offers a third version, stating that configuration management To provide a logical model of the IT infrastructure by identifying, controlling, maintaining, and verifying the versions of all configuration items (CIs) in existence. CIs are all infrastructure components to be controlled – they may be hardware, software or documentation – and contain details describing themselves and their relationships to other components.

Industry organizations, such as Gartner Group, and software vendors such as Encora, also have definitions for configuration management, but the preceding three versions provided us with sufficient input for the team to arrive at a consensus. We all agreed that our definition of configuration management would focus on a proactive approach. We also researched other IT environments similar to our own to learn of their experiences with configuration management.

Configuration management can be applied to any level of the OSI model. It can be applied to the physical networks that connect computing resources, and can also apply to the applications that feed those networks. It can be as broad as managing the security of a networked environment, and as specific as changing a routing table on a network device. Within the enterprise, things are somewhat simpler, but at the same time often less well defined.

Although configuration management can imply a multitude of functions, within the enterprise the focus of configuration management currently is primarily the network itself. But even this can carry its own complexity: It can be as simple as workflow management to automate the various approval processes required for consummating a configuration change, to highly complex automation engines that provide graphical user interfaces that abstract the process of configuration within heterogeneous networks. All of these functions can have a place in the enterprise, but the extent to which an enterprise may want to implement configuration management will be highly dependent on the desired outcome. In most cases, even a little well focused attention to configuration management will yield considerable dividends in improved IT efficiency and reduced costs.

In reviewing these definitions, I believe that we can agree that configuration management should be a proactive IT management concept. Unfortunately, configuration management has been neglected as a comprehensive process by vendors, standards organizations and others in the IT community. The primary reason is that configuration management is exceedingly difficult to implement. Configuration management requires that the organization, from the "top down" gain an understanding of the processes that make up configuration management, and only then implement those pieces that the organization can support, from both a personnel and fiscal standpoint.

To help us determine what pieces of configuration management we could best support, we participated in a very effective brainstorming session referred to as a SWOT analysis. This session enabled us to identify and prioritize the strengths, weaknesses, opportunities and threats (SWOT) associated with configuration management. From this point we proceeded with the development of the process.

4.21.3 Lessons Learned from Initial Efforts

The following are five lessons we learned during the initial phases of implementing a configuration process.

1. Assemble cross-functional team – One of the best approaches we used was to assemble a cross-functional team of key representatives of all of the infrastructure groups. This provided valuable input from a variety of viewpoints and critically needed buy-in across the infrastructure.
2. Agree on a definition of configuration – I found that by identifying several definitions of configuration management from reputable sources led us to a definition that all parties agreed with and understood. It emphasized a proactive approach and resulted in strong buy-in and support from executives and team leads alike.
3. Research what others have done similar to your environment – I sought out other IT environments similar to my own such as Clark County and the Los Angeles Police Department to learn of their experiences with configuration management. This saved us time and effort in learning which products and procedures worked best. For example, trying to find a single product that would address all of our needs was not realistic as others found out. Instead, we scaled our project down in scope and used multiple products linked together.
4. Conduct a SWOT session at the outset – We participated in a professionally-facilitated brainstorming session to identify and prioritize our strengths, weaknesses, opportunities and threats (SWOT) concerning configuration management. This helped us to exploit the strengths we already had such as in-house tools and to mitigate weaknesses such as lack of integration with other processes.
5. Bring in outside professional help when appropriate – We hired a professional consultant who led us through a brainstorming session to assess where we were and the best method to use to get to where we wanted to be. This enabled us to focus on only the parts of our environment most critical to our needs.

In part two of this series I will describe how the LVMPD-ITB initiated efforts to begin designing a configuration process, how the Information Technologies Bureau (ITB) of the LVMPD was currently gathering and managing configuration data, and some of the lessons we learned from these efforts.

References

Schiesser, Rich, IT Systems Management, Prentice Hall, 2002