by Bu Glyn Meek

Prioritization: Systems Management Hierarchy

Oct 15, 20005 mins
IT Governance

Many businesspeople are familiar with Abraham Maslow’s vaunted hierarchy of needs. I propose a similar philosophy in regard to systems management.

Maslow’s primary observation was that some human needs take precedence over others. For example, if you’re thirsty and hungry, your instinct is to satisfy thirst before satisfying hunger. Likewise, if you’re thirsty, but you’re being choked, your instinct is to breathe first, then get a drink of water.

Maslow grouped these needs into five layers: physiological, safety, belonging, self-esteem and self-actualization. His premise is that human needs progress through the hierarchy and, as needs are met at one layer, we seek to satisfy the needs of the next higher level.

What does Maslow have to do with systems management? Quite a lot, actually.

Let’s assume that systems management is less complex than the human psyche. This may seem a stretch for those who deal with the vagaries of mixing four flavors of Unix on 200 servers with 5,000 PCs running Windows 98. But go with me on this.

So instead of five layers, I propose a three-layer hierarchy of systems-management needs: knowledge and control, internal policies and procedures, and continuous improvement.

It’s important to examine how each layer relates to the others. For example, an organization may have satisfied the need for knowledge and control of its resources. Yet without the internal policies and procedures to fully utilize this knowledge, the IT staff may still feel exhausted at the end of each day, and not perceive how their knowledge and control has any positive effect on their company’s core business.

Knowledge and Control

Keeping today’s complex client/server networks running grows progressively more difficult with the proliferation of new hardware and software. Before any IT department attempts to manage its resources, staffers need knowledge of the equipment for which they are responsible.

Systems management faces daily basic questions. How many PCs are in my site? How many have copies of Microsoft Word loaded? How many don’t have virus-scanning software loaded? These are typical knowledge-related issues.

Then there are control-related issues such as calls to IT support. Help requests arrive from across the organization, and the need to track and control them is fundamental.

IT departments should also check for outbreaks of computer viruses, managing the source and proliferation of these through their networks.

Once the fundamentals are covered, there are additional needs to meet depending upon your environment and the size of your budget: software distribution; website monitoring and analysis; backup and restoration (servers and clients); enhanced system monitoring and system performance statistical analysis.

With the basics covered, it’s time to move to the next layer.

Policies and Procedures

IT might know how many PCs its company has or that 90 percent of the virus outbreaks originated from Jane’s PC in accounting, but of what use is this knowledge?

Establishing internal policies and procedures is the next step. It’s important to remember that a systems-management policy designed for a 50,000-node Fortune 500 company, tuned by expensive Big Five consultants and passed down like an Eastern mystic’s teachings may be ill-suited for your organization of 300 PCs and 10 servers. You need practices that are optimized for your environment because implementing a practice designed for another environment may do more harm than good.

When you’re knowledgeable about your operation’s equipment, it’s possible to specify purchasing and support policies to standardize as much as possible. You can standardize on vendors, platforms and applications. One reason Southwest Airlines is so successful is that they standardized on the Boeing 737 airplane for all flight operations, tremendously reducing spare parts inventory, employee training and repair schedules.

Controlling viruses within the organization calls for development of e-mail policies agreeable to everyone—a nontrivial task the first time you introduce them. Procedures for receiving e-mail attachments or FTP downloads are easily implemented and adhered to by all—once you know where files come from and are stored. Procedures for educating users about new viruses increase awareness of associated problems, heading off outbreaks before they start.

With help-desk tools deployed, introducing policies and procedures for problem reporting and tracking significantly speeds up their resolution. Additionally, problem-trend analysis highlights system bottlenecks and can lead to the introduction of procedures to minimize them.

When you’ve satisfied the need for policies and procedures, it’s time to address day-to-day operations.

Continuous Improvement

Your systems management operation now runs smoothly. You are out of fire-fighting mode and are no longer perceived as overhead. Your staff still works long hours, though, and much of what you do is still reactive. What you need now is to increase your value to the corporation and the job satisfaction of your staff. It’s time to introduce continuous improvement components.

Even IT organizations that struggle for knowledge and control plan a certain number of activities. It is only at this layer, where the appropriate controls, knowledge and procedures are in place, that planning becomes meaningful.

When IT organizations contract with support services suppliers, they invariably establish service-level agreements (SLAs) that define performance expectations. Establishing similar SLAs with internal users helps define the IT organization as a true partner in the business. This SLA typically addresses performance, and network and system availability. Additionally, problem-resolution mechanisms are defined, and the help desk becomes proactive in alerting users of problems and resolutions.

In the continuous-improvement layer, planning systems and network capacities based upon current and future requirements is key. Further, an IT organization starts to manage total cost of ownership of PC equipment.

Without the pressure of constant fire fighting, the morale of the IT staff improves, motivation increases and the organization becomes significantly more stable as turnover decreases.

This hierarchical approach to systems management focuses on the current needs of an IT organization with an eye to the future. Knowledge and basic control of systems is the foundation to a stable organization, and it should be the primary focus of IT departments struggling to keep their heads above water.