theartofservice

Most data centers are actually very inefficient. Just the shear number of technologies, services, programming languages, methodologies, applications, and hardware types require a diverse set of skills and experience that require specialized people working on distinct aspects of the IT infrastructure, none of whom have the time and may not have the desire to interact with other members of the whole team.

Add to the mix, the number of concerns that have to be addressed throughout the day. Security, heat, space, bandwidth, capacity, availability, redundancy, backups, migrations, replication, monitoring, and an endless schedule of activities to just maintain the environment for the business. And we have yet to bring in the customer complaints, requests, and expectations. By the end of the day, the data centers become a complex set of systems and applications that have been uniquely designed and implemented into the environment. And the result is a wasteland of confusion and specialized duties.

This isn’t anything new, it’s a old pattern from the mainframe days where entire systems were built for a single function, while another system was built for another function and so on. This was in the time when the interoffice memo was still typed, printed, and physically handed out at a departmental meeting where everyone gathered in single conference room.

The key to an effective data center is simplicity. The technologies today allow systems and applications to be consolidated, integrated, standardized, automated, and modularized in different configurations to make things easier. Here are some simplicity rules for the IT professional:

• Minimize variation – every variation in an environment is one more item that requires attention to ensure that everything works appropriately. Most IT departments have created minimal standards for hardware to limit hardware variation within a specific range.
• Share similar functions – some applications perform the same functions, such as emailing. Unfortunately, these applications typically have different programming required to perform the function. Modularization creates a single set of programming that is used by all applications.
• Focus on priority items, automate the rest – Easier said than done sometimes, automation is useful in those situations when very little to no variation is present and in most cases is reusable for multiple systems.

The first step in building efficient data centers is taking the next step in a sure and confident manner.