Clarifying concepts about a Data Center

I recently had the opportunity to read a very interesting article that clarifies some definitions and solutions related to the world of data centers. It was published in CIO magazine and authored by Michael Bullock, CEO of Transitional Data Services (TDS).

Below is the article I have translated and divided into several parts according to the aspect it covers, for better understanding:

What is a data center?
Known as the server farm or server room, the data center is where most of a company's servers and storage units are located, operated, and managed. There are four main components of a data center:

Blank spaceTypically, this refers to the usable raised floor area and is measured in square meters (it can range from a few hundred to a thousand square meters). For data centers that do not use raised floors, the term "white space" can still be used to indicate usable square meters.

Support infrastructureThis refers to the space and equipment needed to support data center operations, including power transformers, uninterruptible power supplies (UPS), generators, air conditioning units (CRAC), remote transmission units (RTUs), chillers, air distribution systems, etc. In a high-density, Tier 3 data center (i.e., with facilities for concurrent maintenance), this support infrastructure can consume 4-6 times more space than the white space and must be taken into account in data center planning.

IT TeamsThis includes the racks, cabling, servers, storage, management systems, and network equipment necessary to provide computing services to the organization.

Operations: Operations staff ensure that systems (both IT and infrastructure) are properly managed, maintained, updated, and repaired when necessary. In most companies, there is a division of responsibilities between the IT technical operations group and the staff responsible for supporting facilities systems.

How to manage a data center?

Operating a data center with maximum efficiency and reliability requires the combined efforts of Facilities and IT.

IT systems: Servers, storage devices, and network equipment must be maintained and updated. This includes things like operating systems, security patches, applications, and system resources (memory, storage, and CPU).

Facilities Infrastructure: All support systems in a data center face heavy loads and must be properly maintained to continue operating successfully. These systems include cooling, humidification, air treatment, power distribution, backup power generation, and much more.

MonitoringWhen a device, connection, or application fails, it can cause mission-critical operations to collapse. Sometimes, a system failure triggers cascading failures of applications in other systems that rely on the data or services of the failed unit. For example, multiple systems, such as inventory control, credit card processing, accounting, and more, are involved in a process as complex as e-commerce payments. A failure in one compromises all the others. Furthermore, modern applications typically have a high degree of device and connection interdependence. Ensuring maximum uptime requires 24/7 monitoring of the key applications, systems, and connections involved in the various workflows across an entire enterprise.

Building Management System (BMS): For larger data centers, the building management system (BMS) will allow constant and centralized monitoring of the facilities, including temperature, humidity, power, and cooling.

IT and data center facilities management is often outsourced to third-party companies that specialize in monitoring, maintaining, and repairing systems and facilities on a shared services basis.

What is a green data center?

A green data center It is the one that can operate with maximum energy efficiency and minimum environmental impact. This includes mechanical, lighting, electrical, and IT equipment (servers, storage, network, etc.). Within companies, the focus of the  green data center This is primarily due to a desire to reduce the enormous electricity costs associated with operating a data center. In other words, going green is recognized as a way to significantly reduce IT infrastructure operating expenses.

Interest in the green data centers It is also being promoted by the U.S. federal government. In 2006, Congress passed Public Law 109-431 asking the EPA to: “analyze the rapid growth and energy consumption of data centers by the Federal Government and private companies.”.

In response, the EPA developed a comprehensive report analyzing current trends in energy use and energy costs for data centers and servers in the U.S. and highlighted existing and emerging opportunities for improving energy efficiency. Recommendations were also made for leveraging these energy efficiency opportunities nationwide through information and incentive-based programs.

According to the EPA report, the two largest electricity consumers in the data center are the following:

• Support infrastructure - 50% of the total

• General servers – 34% of the total

Since then, significant progress has been made in improving server efficiency. High-density and blade storage servers now offer much more computing power per watt of energy. Server virtualization is enabling organizations to reduce the total number of servers they support, and the introduction of Energy Star servers has combined to offer many options for both the public and private sectors to reduce the overall electricity consumption of servers.

Of course, the greatest opportunity for additional savings lies in the supporting infrastructure of the data center facility itself. According to the EPA, most data centers consume 1,001 to 3,001 TP3T of additional energy for the support systems they use for their core IT operations. Through a combination of best practices and migrating to facility upgrades with a rapid return on investment (such as ultrasonic humidification and airflow control), this overhead can be reduced to 301 TP3T of the IT load.

Who are the main stakeholders interested in the data center?

While the data center must provide the necessary resources for end users and enterprise applications, the provisioning and operation of a data center is divided (sometimes by force) between IT, facilities, and finance, each with its own unique viewpoint and responsibilities.

YOU: The IT manager is responsible for the company's IT group, making decisions about which systems and applications are needed to support business operations. IT will directly manage the aspects of the data center that are directly related to IT systems, while Facilities will handle the provision of power, air conditioning, access, and physical space within the data center.

Facilities: The Facilities team is generally responsible for the physical space—for its provision, operation, and maintenance—along with other building assets owned by the company. The Facilities group, in general, will have a good understanding of the overall efficiency of the data center and will have a grasp of the IT data load and total energy consumption.

Finance: The finance group will be responsible for aligning short-term and long-term capital expenditures (CAPEX) to acquire or improve physical assets and operating expenditures (OPEX) to execute with the set of corporate financial operations (balance sheet and cash flow).

Perhaps the greatest challenge these three groups face is that, by its very nature, a data center will rarely operate at or near its defined optimum range. With a typical lifecycle of 10 years (or perhaps more), it is essential that the data center design be flexible enough to support increasing power densities and varying occupancy levels over a significant period. This inherent flexibility must be applied to power, cooling, space, and network connectivity. When a data center is approaching its power, cooling, and space limits, the organization faces the need to optimize its existing facilities, expand them, or establish new ones.

What options are available when you're reaching the limits of energy, space, and cooling?

Optimize: The quickest way to address this problem and increase available power, space, and cooling capacity is by optimizing an existing facility. The greatest optimization gains can be achieved by reducing the overall server power load (through virtualization) and improving facility efficiency. For example, up to 70% of the power required to cool and humidify the data center environment can be conserved using currently available technologies such as outdoor air economizers, ultrasonic humidification, high-efficiency transformers, and variable frequency drives (VFDs). Using these techniques in conjunction with new high-density IT systems will allow many facilities to increase their IT capacity while simultaneously reducing facility overhead.

Move: If your existing data center can no longer be modified to support high-density solutions, there may be no other option but to move to a new space. This process will likely begin with a needs assessment and site selection process and conclude with either a remodel of your existing facility or a relocation to a new building.

Outsourcing: Besides proceeding with their new own facility, there are two other options worth considering:

• Colocation: This means moving your data center to a shared facility managed by a suitable service provider. There is a wide range of business models under which these services can be provided (including full business responsibility). It's important to ensure that the terms of the agreement match your short- and long-term needs and (always) consider the flexibility you require for your data center to evolve throughout its lifespan.

• Cloud computing: The practice of leveraging shared computing and storage resources – and not just the physical infrastructure of a colocation service provider – has been growing rapidly for some niche application-based products.

What are some data center metrics and benchmarks, and where can I find them?

PUE (Power Usage Effectiveness)Created by members of The Green Grid, PUE is a metric used to determine the energy efficiency of a data center. The data center's PUE is calculated by dividing the amount of incoming power by the power used by the IT infrastructure. Expressed as a ratio, efficiency improves as it approaches 1. Data center PUE typically ranges from around 1.3 (good) to 3.0 (bad), with an average of 2.5 (not so good).

DCiE (Data Center Infrastructure Efficiency)Created by members of The Green Grid, DCiE is another metric used to determine the energy efficiency of a data center, and it is the reciprocal of PUE. It is expressed as a percentage and is calculated by dividing the power of the IT equipment by the total power of the facility. Efficiency improves as DCiE approaches 100%. A data center's DCiE typically ranges from 33% (poor) to 77% (good), with an average DCiE of 40% (not so good).

LEED CertifiedDeveloped by the U.S. Green Building Council (USGBC), LEED is an internationally recognized green building certification system. It provides third-party verification that a building was designed and constructed using strategies intended to improve performance across all metrics: energy savings, water efficiency, CO2 emissions reduction, indoor environmental quality, resource management, and sensitivity to its overall environmental impact. For more information about LEED, go towww.usgbc.org.

The Green GridThe Green Grid is a global consortium of non-profit organizations, government agencies, and educational institutions dedicated to improving energy efficiency in data centers and enterprise computing ecosystems. The Green Grid does not endorse specific product or solution vendors and instead seeks to provide the entire industry with recommendations on best practices, metrics, and technologies that improve the overall energy efficiency of data centers. For more information about The Green Grid, go to [link to Green Grid website]. http://www.thegreengrid.org.

Telecommunications Industry Association (TIA): The TIA is the leading trade association representing the information and communications technology (ICT) industries worldwide. It helps develop standards, gives ICT a voice in government, provides market intelligence, certification, and promotes business opportunities and global environmental compliance. Supported by its 600 members, the TIA enhances the business environment for companies involved in telecommunications, broadband, mobile wireless, information technology, networking, cable, satellite, unified communications, emergency communications, and technology greening. The TIA is accredited by ANSI.

TIA-942: Published in 2005, the Telecommunications Infrastructure Standards for Data Centers was the first standard to specifically address data center infrastructure and was intended for use by data center designers at the beginning of the development process. TIA-942 covers:

• Site space and design

• The cabling infrastructure

• Reliability in levels or tiers

• Environmental considerations

Reliability levels – The TIA-942 standard has been adopted by ANSI for reliability levels based on its usefulness in the overall evaluation of redundancy and availability in a data center design.

Tier (Rating) 1 Basic – no redundant components (N): 99.671% availability

• Sensitive to disruptions to planned and unplanned activities

• Single path for feeding and cooling

• It must be completely switched off to perform preventive maintenance

• Annual downtime is 28.8 hours

Tier (Rating) 2 – Redundant components (limited N+1): 99.741% availability

• It is less susceptible to disruptions from planned and unplanned activities

• Single path for power and cooling includes redundant components (N+1)

• Includes a raised floor, UPS and generators

• Annual downtime is 22.0 hours

Tier (Rating) 3 – Concurrent maintenance (N +1): 99.982% availability

• It allows planned activities (e.g., scheduled preventive maintenance) without interrupting hardware operation (unplanned events can still cause interruption).

• Multiple power and cooling paths (one active path), redundant components (N+1)

• Annual downtime is 1.6 hours

Tier (Rating) 4 – Fault Tolerant (2n +1): 99.995% availability

• The planned activities do not interrupt critical operations and can sustain, at least in the worst-case scenario, an unplanned event without impacting the critical workload

• Multiple active power and cooling paths

• Annual downtime is 0.4 hours

Due to the duplication of infrastructure (and space) in a Tier 3 facility, a Tier 4 facility will have a much higher cost to build and operate. Consequently, many organizations prefer to operate at the more economical Tier 3, as it achieves a reasonable balance between CAPEX, OPEX, and availability.

Uptime Institute: It is a non-profit organization created to achieve consistency in the data center industry. The Uptime Institute offers education, publications, consulting, research, and conference series for the enterprise data center industry.

What data center technologies should I consider?

Alternative energy sources—solar, wind, and hydropower—show great potential for generating electricity in an environmentally friendly way. Nuclear and hydropower also show great potential for green energy. However, the biggest challenge when it comes to using alternative energy for your data center applications is the need for a constant supply at high service levels. If you use alternative energy but still have to buy from the local power company when you encounter peak loads, many of the economic benefits you're getting from the alternative energy source will quickly disappear. New storage mechanisms have been developed to capture and store excess capacity so it can be accessed when needed. Soon, alternative energy sources will play a much larger role in the data center than they do today. Air and water storage systems show great potential as environmentally friendly energy storage options.

Ambient Return: This is a system in which air returns to the air conditioning unit naturally, without being directed. This method is ineffective in some applications, as it is prone to mixing of hot and cold air, and to stagnation caused by static pressure, among other problems.

Chiller-based cooling: It is a type of cooling in which chilled water is used to dissipate heat in the CRAC unit (instead of glycol or refrigerant). The heat exchanger in a chiller-based system can be air or chilled water. Chiller-based systems feature CRAC units with greater cooling capacity than DX-based systems. In addition to eliminating the DX limitation of approximately -4°C spread between the outlet and inlet, the cooling system can be dynamically adjusted based on the load.

Chimney effect: Just like a chimney in a home, it takes advantage of air pressure differences to direct the airflow. This has led to a common design where cool air is drawn in from below a raised floor, then sent to the data center and directed as warm air at the top through the chimney. This design creates a very efficient circulation of fresh air and minimizes air mixing.

Cloud computing: This is a computing model that is dynamically scalable through virtualized resources provided as a service over the internet. In this model, the client doesn't need to worry about the technical details of the remote resources. (That's why it's often represented as a cloud in system diagrams.) There are many different types of cloud computing options with variations in security, backup, control, compliance, and quality of service that should be carefully researched to ensure their use won't put the organization at risk.

Cogeneration: This involves using an engine (typically diesel or natural gas-based) to generate electricity and useful heat simultaneously. The heat emitted by the engine in a data center can be used by an absorption chiller (a type of chiller that converts heat energy into cooling) to provide cooling benefits in addition to electricity. Furthermore, excess electricity generated by the system can be sold back to the grid to cover costs. In practice, the effective return on investment for cogeneration depends largely on the difference between the cost of electricity and fuel. This cogeneration alternative will also contribute to a substantial increase in CO2 emissions.₂ of the facilities. This goes against the trend towards ecological solutions.

Colocation: It is one of several business models in data centers provided by another company. In the colocation option, data centers for multiple organizations can be housed in the same common facility, sharing electrical power, cooling infrastructure, and facilities management. Colocation differs from a dedicated hosting provider in that the client owns their own IT systems and has greater flexibility in which systems and applications reside in their data center. The lines are blurred between the various outsourcing models, with variations in rights, responsibilities, and risks. For this reason, when evaluating new facilities, it is important to ensure that the business terms are properly aligned with your long-term space requirements.

Containers: The idea behind a containerized data center is that all power, cooling, space, and connectivity can be incrementally provisioned using standard-sized shipping containers. These containers can be deployed off-site to expand data center capacity or installed in a warehouse-like environment. One benefit of containerized data center solutions is their ability to enable rapid deployment and their integration and configuration to support very high power densities. Containers have been adopted for use in cloud services by Google and Microsoft. Potential disadvantages of containers include their high cost, their tendency to be homogeneous (designed for specific brands and models of systems), and their requirement for self-contained operation (the container must remain sealed to operate within specifications).

CRAC (Computer Room Air Conditioner): It is a specialized air conditioning unit for data center applications that can increase air humidity to maintain the appropriate humidity level required by electronic systems.