ITSM & Teamwork Run Data Centers Better

In the last few years, standards for data center technology have been refined and improved to the point where data center managers target 99.999% “uptime” for the servers and storage needed to run enterprise systems.

Such improvement is expensive and, with humans involved, ever-increasing complexity of data center technology pushes up the rate of human error. This puts large technology investments at the mercy of common team dynamics and the staff member with the lowest level of knowledge.

On the technical side, since 1995, the Uptime Institute has been refining their description of what constitutes a reliable data center. Its Tier Performance Standards are an objective basis for comparing the capabilities of particular design topologies.

For the most part, these standards concern the redundancy, maintainability, and fault tolerance of critical data center infrastructure components, like power distribution and cooling systems. From these standards, the Uptime Institute correlates levels of reliability that the data center can provide, i.e., “Five Nines” (or 99.999% uptime).

While this is all well and good but the Institute does not define standards for human interaction and team dynamics relating to the performance of data centers.

IT vs. Facilities

The challenges in maintaining reliability within the data center are exacerbated by the division of teams that run it. Generally accepted rule such as COBIT specify the facilities department—typically the group that handles the buildings and space in which the data center exists—as the organization responsible for the data center’s physical space, however, the IT department, with its focus on engineering disciplines, is where the knowledge exists about the equipment.

The culture difference between teams, as well as the attention paid by upper management can cause separation and conflict. Having teams work together for the sake of efficiency and quality can be difficult with over-arching culture clash established in an organization.

The objectives of the teams can be in conflict as well. Incentives have been set up for IT, fueled by government regulations like Sarbanes-Oxley, to keep servers running at all costs, where facilities is concerned with the integrity of the building and its assets.

If a fire breaks out in the data center, the IT guys may be trying to determine which specific servers can be shut down to put out the fire, whereas the facilities guys will be looking to shut down all power to the floor.

Added to this are management’s objectives, which center on more professional and robust management of the services provided by IT, in the form if IT Service Management (ITSM).

Enter ITSM

As the next level of proactive management of IT, ITSM is a set of services that achieve and align IT delivery with business goals. It includes the definition and management of associated service level agreements, maintenance requirements and, in some cases, helps define services which directly impact the level of revenue for the business.

Businesses that want ITSM are becoming increasingly dependent on IT to deliver expansive data storage and computing services, from typical finance and communications support, and critical manufacturing applications, to systems used directly by customers, as in the case of online entertainment and content providers.

With strong management pressure, IT leadership is in a bind to provide more and more services within the data center infrastructure. This leaves some in IT exasperated.

The answers to these problems may be found in the workings of a wide variety of other complex organizations, where strict process is defined to handle the complexity. Usually, data-center process management revolves around change, as changes are the number one cause of unexpected outage in data centers.

Understanding how changes are made to the data center, and how the change will impact the systems running in the data center, are key ingredients to keeping high levels of reliability and preserving uptime.

Having a detailed understanding of what equipment exists in the data center infrastructure is also an important part of teamwork, as this promotes communication among the members of the team. Discussion about problems, and determining the best solutions, should not be encumbered by problems in getting and communicating information.

Within the data center, this is in the realm of configuration management, where the current state of the infrastructure is well documented and understood. Understanding what exists brings greater understanding to those who are trying to fix the present and to plan out the future.

Team integration may also be very important to achieving the future goals of the data center. The emergence of the data center team where goals, objectives and incentives are properly aligned with the goals of the business, is a very promising trend, and one that aligns with ITSM.

Sean Nicholson is director of Product Management at Aperture Technologies, a provider of enterprise-class data center management solutions.