As a definition, data centre infrastructure management (DCIM) refers to convergence of IT and building facilities functions within an organization.

As an objective, DCIM focuses on providing administrators with a holistic view of a data centre’s performance so that energy, equipment and floor space are utilized as efficiently as possible.

Why so much importance is accorded to DCIM?

The answer is straightforward.

In the present scenario, infrastructure management and system monitoring go hand in hand. As demands for colocation and managed data centre are increasing, the need for efficient performance of critical data centre environments is becoming a primary concern.
Today data centre management has typically two cornerstones.

1. Managing the IT and point solutions

2. Managing energy, cooling and other facilities.

"It is pointless to go entirely with what suppliers and equipment providers guarantee with regards to performance", says a CIO of an IT company. "It is imperative that we must gain inclusive control over entire data centre in real time".

The CIO is entirely on target with his observation.

For data centre managers it is a gigantic task to coalesce all the silos of information they are receiving and also deliver added value, enhance efficiencies and affect cost savings.

The mantra – prevention is better than cure applies so aptly to data centre management.

What is the point in summoning a disaster recovery plan when proper prevention can preempt it in the first place?

However, energy saving and capital expenditure reduction are just two of the issues facing data centre managers. The challenges also include the following:

• Improving availability

• Enhancing operational transparency

• Guaranteeing  a better recovery should any disaster occur 

So what is the solution?

A comprehensive data centre infrastructure management can be the answer. 

DCIM tools must integrate diverse components of IT systems, building systems and energy management by:
 
• Monitoring consumption in real time

• Evaluating power usage effectiveness

Not surprisingly perhaps, a study carried out on several data centre buildings came to the conclusion that data centres can be exponentially more energy intensive as conventional office buildings.

This makes  it all the more important to focus on energy efficiency.

The study also offered guidelines or best practice technologies that can greatly improve energy efficiency in data centres.
Here is a list of some of the recommendations.
 
Optimizing air management

With the computing power hitting new highs, it is not surprising that data centers are beginning to experience unexpected high heat loads. In such a scenario, all data centre regardless of their sizes must focus of effective air distribution. 

Any engineering expert will tell you that air distribution has a significant bearing on energy efficiency and equipment reliability.
 
Improvement in air management can be achieved by finding solutions to the following:

• Incorrect location of computer room A/C units

• Inadequate ceiling height

• Air blockages in ducts and piping

• Poor airflow through racks having IT equipment due to configuration issues/ restrictions in flow

It was also observed that data centres that have installed centralized air handler systems perform better than those having conventional multiple distributed unit systems.

Right sizing of data centre

Typically, most data centres are sized taking into consideration future power needs. Therefore it is a challenge for designers and engineers to size electrical and mechanical systems so that they will operate efficiently even when overall loading is much below design.
 
Leveraging "free cooling"

It is observed that data centre equipment cooling loads do not vary much over the entire period of the year. As such, data centres can capitalize of "free cooling" available via evaporative cooling through water side economizers.
 
Evaluate the possibility of on-site generation of power

This may sound tricky, but a proper techno-economic evaluation of the possibility of installing an on-site power generating unit is not out of place. It can prove to be a good alternative to grid power, and moreover the waste heat generated by such units can be used to meet neighboring heating needs. 

Yes, it is true CAPEX has to be taken into account, but for larger data centres the long run positives especially with regard to providing high degree of reliability to customers can make an on-site power unit a sensible choice.

Improvements in miscellaneous operations and maintenance processes

It goes without saying that by putting in place best practices via better technologies and operational procedures can go a long way in improving the overall performance of data centres. 

The following points can be kept in mind:

• Put in place a strong sustained energy management program that is integrated with processes and other functions. 

• Adopt best engineering practices from the conception and design stage itself. 

• Include in the data centre facility, tools that perform integrated monitoring, measuring and control.

• Include periodic training for the staff in energy management.

Implementing best practices in data centres is a continuously evolving process. This is true especially for large and medium data centres that are high energy intensive. But if judiciously executed, data centres can surely see a positive return on investment.