ID Technologies Logo
CACI Acquires ID Technologies – Learn More

Data Center Infrastructure 101.

A helpful guide to understanding the basics - and some of the not-so-basics - of data center infrastructure.

What is a data center?

A data center is a network of servers and support components assembled to power applications and process, store, and securely distribute often massive amounts of data.

Applications supported by data centers include email and file sharing, customer relationship management tools (CRMs), enterprise resource planning (ERP), virtual desktops, big data, artificial intelligence (AI), and machine learning.

Traditionally, "data center" has referred to the physical components comprising a data center. The physical facilities housing computing power, storage infrastructure, security systems, and networking abilities.

Data center facilities can vary in size from underwater cabinets to multi-building campuses spread across geographically optimal locations.

However, advancements in cloud technology have allowed the majority of modern data center operations to exist off-premises, in the virtualized environments of shared networks.

The modern data center must be able to connect secure and reliable on-premises systems with multiple public and private cloud and edge infrastructures. Data centers support critical applications and data management for organizations while providing secure, shared network accessibility to verified users.

The use of data centers is booming.

The Basics of Data Center Infrastructure.

Data center infrastructure or "DCI" refers to a complimentary assembly of physical data center components designed to optimize the performance of applications or management of data.

DCI components include supporting hardware, such as server racks, cooling fans, and support structures, as well as software resources, like data center infrastructure management systems. 

Because most applications no longer reside within a single, physical data center, the term data center now commonly refers to the department responsible for data center systems regardless of location. While data center infrastructure still generally refers to physical components.

The Importance of Connectivity.

The modern data center is increasingly capable of utilizing virtual networks to pool physical infrastructures.

Building optimized virtualized environments on the foundation of optimized physical environments.

Data centers can utilize cloud, multi-cloud, and edge solution environments to share assignments based on best fit for the task. The connected networks of different physical data centers create a much wider-ranging and diverse pool of resources to draw from than what could traditionally be achieved by a stand-alone data center.

Integrated network services and established data center interoperability standards now allow disparate systems to function within a single environment.

This reliance on connectivity means data centers must be able to communicate immediately, effectively, and reliably across multiple physical and virtual environments. Safe and reliable accessibility has never been as important.

Learn More on Virtual Desktop Infrastructures (VDI)!

Data Center Design: why certain components are chosen.

Data center infrastructure design aims to assemble the best possible collection of both core and support components toward one or more desired data center operations.

The assembly of components can be tailored to a single, specific task, or multiple simultaneous operations.

Additional industry-wide service level agreements help ensure users achieve desired outcomes.

Established companies, like IDTec, can draw on an expansive pool of tried-and-proven component solutions to quickly design and assemble the perfect data center to meet your organization's specific requirements.

Learn More on Hyperconverged Infrastructure (HCI)!

Core Infrastructure.

It’s all important. But this stuff’s really important.

These components perform the core operations of a facility and are often the most well-known parts of a data center. Core components include servers, computers, networking equipment (routers and switches), security (firewalls and biometric security systems), storage (storage area networks [SAN]), and backup/tape storage, as well as data center management software and applications.

Core physical components include non-computing resources such as physical server racks, chassis, and cabling. This data center hardware houses and connects the core computing equipment found in a standard data center.

SERVERS

The beating heart. The roaring engine. The Oz behind the curtain.

You get it.

Servers are the things everything else in a data center is built to support.

Servers are high-powered computers built to process, store, and manage data. Within a data center servers manage and execute applications for devices and systems.

Common Types of Server Hardware:

  • Tower Servers
  • Rack Servers
  • Blade Servers
  • Mainframes 

STORAGE

The librarians, the bookkeepers, maybe even the nerds of a data center.

Data center storage infrastructure is the combined hardware, software, and processes that monitor and manage data storage.

On-site storage infrastructure components include hard disk drives, tape drives, solid state drive (SSD) flash arrays, direct-attached storage (DAS) devices, and redundant array of independent disks (RAID) devices, as well as technologies like storage area networks (SAN) and network attached storage (NAS).

Storage infrastructure also includes policies and procedures governing data storage and retrieval. These policies encompass security, collection, access control, and data availability.

Learn More on Data Center Storage!

NETWORK

The dealmakers. The agents. The social glue. Making sure everyone knows everyone and no one’s left out.

Data center network infrastructure is a constellation of networking resources that provides connectivity between data center components, users, and internal and external resources, to support the storage and processing of applications and data.

The connectivity of data center networks allows access to enterprise applications nearly anywhere in the world. Networking resources include cabling, routers, switches, firewalls, load balancing, analytics, and much more.

Effective data center networking can manage high volumes of traffic without compromising performance. Data center networking is essential to the integration of different data center hardware systems.

Data center networks are most often two or three-tiered. A three-tiered network features core switches, distribution layer switches, and access switches. Two-tiered networks feature spine switches and leaf switches.

Technological advancements including hyperscale and software-defined networking have brought unlimited, frictionless scalability to on premises networks.

Support Infrastructure.

The support infrastructure of a data center is the components dedicated to enabling other components to best perform their own operations.

A data center (and its applications) is considered available and in uptime when it is operational. The goal of a data center support infrastructure is to maximize a data center’s availability.

Effective data center support infrastructure offers unfettered operational environments for core systems and applications.

Different types of physical support components have been designed to address different data center needs.

Powering a data center.

Arguably the most important ability of a data center is its ability to turn on. A data center’s power components make sure it can turn on and stay on without interruption.

This means providing and maintaining both primary and backup power supplies as well as an electrical infrastructure capable of powering the entirety of servers and equipment within the data center.

A clean, reliable power source is of utmost importance in keeping equipment running and maintaining a data center’s uptime. Most DCs tap into municipal power grids, with failover or backup sources available if necessary.

Uninterruptible Power Supplies (UPS), like batteries or diesel generators, offer further redundancy in backup power.

Examples of data center power components:

  • Electrical plugs for servers
  • Electrical wiring to connect individual internal components.
  • Electrical wiring to connect data center to the municipal power grid.
  • Fuel
  • Solar panels
  • Turbines - HVAC, wind

Cooling a Data Center

The physical components of an operational data center tend to generate significant heat.

Having systems in place to manage temperatures and mitigate this heat is essential. Cooling components are data center hardware and software designed to monitor and manage the temperature and climate within a data center. These components aim to create an optimal environment for the operation of data center hardware.

Fans, HVAC, and pipes filled with cold water running alongside warm infrastructure, all help to ensure servers and other computing infrastructure hardware remains at optimal temperature and doesn’t overheat.
Structural components, like hot and cold aisles, can be designed and built into a data center’s architecture to draw heat away from active components and pump in cold air.

Fire suppression systems in case of critical failure are an essential safety component.

Building a Data Center

Structural components of DCI are an underappreciated method of optimizing performance.

However, the challenge with these physical solutions is that many cannot be added once a data center has been built. Which is why the design stage is so important when it comes to structural solutions.

Raised floor architecture creates a gap between the true floor and another floor containing additional IT equipment. The gaps house electrical wiring, cabling, cooling, and equipment, and make it easier for technicians to access connectivity and cooling infrastructure.

Another important structural element is the ability of a data center to physically withstand major acts of nature. Steal-reinforced concrete is just one of many provisions against things like floods, hurricanes, earthquakes, tornadoes, blizzards, and heat waves.

Securing a Data Center

Because of the sensitive nature of so much of the data and applications contained within data centers, access to floors and buildings housing centers is typically guarded with locked doors and robust security infrastructures including security personnel.

Server rooms are often locked and isolated with additionally limited access compared to the rest of the facilities.

Five Different Types of Data Centers

Evolution of data center infrastructure has caused the growth and subsequent classification of several types of data center facilities. These classifications depend on the number of organizations that own the data center, how the data center fits into the topology of other data centers, what technologies are used for computing and storage, and energy efficiency.

1. Enterprise Data Centers

Traditionally organized facilities directly owned and operated by a single organization. Typically constructed and operated by a single organization towards its own internal purposes and optimized for end users.

Enterprise data centers are usually located on-site and housed on corporate campuses, allowing in-house teams to oversee maintenance, IT deployments, hardware upgrades, and network monitoring.

On-premises data centers often require expensive real estate and resource investment. This type of data center is best suited for applications that can’t be moved to the cloud due to security, compliance, and other related reasons.

2. Collection/Colocation ("CoLo") Data Centers

Collection data centers are shared spaces in which organizations and companies can rent space for servers and other hardware off-premises. While a portion of the data center infrastructure is shared, organizations are still responsible for the provision and management of hardware components including servers, storage, and firewalls.

A major benefit to this type of data center arrangement is the offloading of responsibility for building, power, support, bandwidth, and physical security.

3. Managed Services Data Centers

Managed data centers are, as the name might suggest, managed by a 3rd party on behalf of an organization. Under a managed data center arrangement, a company or organization leases the data center’s equipment and infrastructure rather than buying it. A 3rd party managed service provider oversees the hardware and facility, offering data storage, computing, and other services to customers directly.

4. Cloud Data Centers

This type of data center has become increasingly popular in recent years. A cloud data center represents an off-premises facility that is internet accessible to an organization and in which the customer has no responsibility for maintaining the necessary infrastructure. Public clouds are collections of data centers and applications hosted in a cloud that use data center resources from the cloud provider.

A multicloud data center refers to cloud services provided to a single user or organization from two or more public cloud providers. This terminology describes specific architectures allowing an application to use the same service model across multiple cloud providers, such as combining on-premises data centers with colocation facilities.

One of the most advantageous features of cloud data centers is their ability to be provisioned or scaled simply and instantaneously. Infrastructure-as-a-Service (IaaS) offerings can spin up whole systems on demand. While Platform-as-a-Service (PaaS) and container technologies offer instant availability towards the creation of new, needed applications.

Data and applications are currently hosted by a wide variety of private and public cloud service providers such as Amazon Web Services (AWS), Microsoft (Azure), and IBM Cloud.

Distributed data centers are sometimes offered to customers with the help of 3rd party managed service providers.

Dive Deeper on Cloud Data Centers and Hybrid Cloud Data Centers!

5. Edge Data Centers

A more recent development, edge data centers are small, decentralized facilities that exist and operate close to the edge of a network. These data centers, distributed among remote nodes, typically house equipment offering compute and storage in locations closer to the end user. This virtual proximity can reduce latency, optimize bandwidth, and improve performance.

Backup systems are often placed in edge data centers, providing users with better access to data.

Learn More on Edge Data Centers!

Important Data Center Standards and Regulations

Data center infrastructure standards are based on levels of potential risk of service interruption. These industry standards exist to assist in the design, construction, and maintenance of data center facilities and infrastructures. The goal DCI standard is to ensure data is both secure and highly available.

Foundational service level agreements (SLAs) establish expected service reliability within a calendar year. Data centers can deploy increased redundant resources to increase reliability and reduce the chance of downtime. The time a data center is functional and fulfilling its service agreements is referred to as uptime and is expressed as a percentage.

The most widely adopted DCI standards were created around ANSI/TIA-942-ready certification. These certifications correspond to compliance with one of four tiers rated for redundancy and fault tolerance levels.

Tier 1: Basic Site Infrastructure/Basic Capacity

Tier 1 data centers feature single-capacity components and a single, non-redundant distribution path, offering limited protection against physical events.

This basic capacity must include a UPS and suffer no more than 29 hours of service interruption over a calendar year (99.671% uptime).

Tier 2: Redundant-Capacity Component Site Infrastructure

Tier 2 data centers feature improved protection against physical events through redundant-capacity components and a single, non-redundant distribution path.

These data centers often add redundant power and cooling elements and must maintain no more than 22 hours of service interruption over a calendar year (99.741% uptime).

Tier 3: Concurrently Maintainable Site Infrastructure

Tier 3 data centers can protect against virtually any physical event by providing redundant-capacity components and multiple independent distribution paths.

Any individual component of this tier of the data center can be removed or replaced without disrupting services to end users or affecting production. These data centers must maintain no more than 1.6 hours of service interruption over a calendar year (99.982%).

Tier 4: Fault-Tolerant Site Infrastructure

Tier 4 data centers offer the highest level of fault tolerance and redundancy through redundant-capacity components and multiple independent distribution paths that enable concurrent maintainability. Preventing faults within an installation from causing downtime.

Production capacity of these types of data centers is insulated from failure, and faults within installations will not result in downtime.

This infrastructure must maintain no more than 26.3 minutes of service interruption over a calendar year (99.995%).

Making Sure Everything's Secure.

Due to the business-critical data and applications housed within data centers, robust security mechanisms are integral to effective data center design. The primary goal of DC security components is to protect the performance and integrity of a data center’s core components.

Security protocols such as firewalls, data access controls, IPS, WAF, and Web Application and API Protection (WAAP) systems must scale to meet the demands of data center networks without compromising effectiveness.

Thorough zero-trust analyses must be incorporated into any DC design. Network security appliances, such as firewalls and intrusion protection systems, offer an initial level of security. While application delivery assurance mechanisms strengthen application resiliency, and availability and maintain application performance via automatic failover and load balancing.

Data Center Infrastructure Management (DCIM) Solutions

Data center infrastructure management refers to the oversight and operation of combined IT and building facility functions within a data center.

DCIM is intended to provide administrators with the necessary information to optimize every aspect of a data center’s performance. These systems utilize discovery, monitoring, reporting, and visualization to control data center operations effectively.

Elements of DCIM can be performed by in-house teams, DCIM software, managed service providers, 3rd party maintenance, and any combination of these.

DCIM - the boss.

DCIM is often performed by a team of data center operations managers of similarly titled employees.

These managers are responsible for the maintenance and upkeep of all infrastructure and often design and implement procedures and practices to ensure infrastructure is well-maintained, available, and consistently upgraded when necessary.

In addition to a core DCIM team, many organizations employ specialists capable of ensuring specific infrastructure is maintained and updated regularly. For example, a dedicated HVAC team working as a contractor to ensure and maintain optimal temperature conditions within a data center.

DCIM Tools - tangible ways to make things easier.

DCIM tools offer the ability to monitor, measure, manage, and control all aspects of data center performance. This includes the optimization of energy consumption of all IT-related equipment and infrastructure components.

These tools help administrators locate, identify, and resolve active or potential conflicts in relationships between data center infrastructures and their IT systems. Energy monitoring sensors can measure and manage energy consumption while at the same time analyzing power usage effectiveness and cooling system energy efficiency.

DCIM Software - let the computers handle the tough stuff.

DCIM software represents one of the most influential and essential tools available to DCI managers and administrators. DCIM software measures, monitors, and manages a data center's IT equipment and supporting infrastructure, enabling operators to run efficient operations while improving infrastructure design planning.

Optimal DCIM software is easily integrated with existing management software to better track automated and integrated data center workflows. DCIM software also performs significant data analysis, including the collection of real-time data, such as hardware metrics, to help mitigate outages and other incidents.

DCIM software can be hosted on-premises or in the cloud.

Some of the benefits of DCIM...

Effective DCIM represents a significant performance advantage when utilized alongside your data center infrastructure. This can include:

Increased Uptime: Organizations can quickly and effectively determine data center health and the need for equipment maintenance and replacement through monitoring equipment and constructing reports on gathered data.

Capacity Management: DCIM software can help organizations model and allocate space and resources to support new hardware and equipment and optimally manage their power chain.

Proactive Incident Management: Real-time facility and operations monitoring can warn administrators and managers of likely and possible issues and complications before they would be noticed.

Energy Management: DCIM software can monitor, suggest, and implement best practices for energy consumption for all IT-related equipment, reducing energy usage and costs.

Improved Productivity: DCIM software enables IT departments and management teams to monitor multiple locations simultaneously and remotely. Automated alerts allow these teams to track physical and logical components and warn of active and potential failures.

Learn More on Data Center Infrastructure Management (DCIM)!

Effective DCIM means optimization of your data center investments.

The rapid advancement of data center technologies has made effective management of data center operations nearly impossible without software assistance.

While the growing variety of applications and means solutions must be easily and quickly customized and deployed.

DCIM software is already critical for organizations and businesses across the globe. And potential benefits to performance appear limitless.

A failure to utilize the best DCIM for your organization's needs can mean your losing money.

Learn More on Data Center Optimization!

IDT Can Help You Build the Optimal Data Center Solution to Fit Your Unique Needs.

IDTec is a team of experts dedicated to creating customized data center solutions to help you meet your specified objectives.

Over a 25-year IT partnership with the U.S. Government, we've learned and earned our place as an industry leader in providing secure and reliable operating environments.

Our diverse library of proven, CSfC solutions means a deployable, secure environment is just a conversation away.

Ready to Take Action?

Share Your Email to Find Out How IDT Can Match You With The Optimal Data Center For Your Organization's Needs.
START HERE
ID Technologies Logo
We have a clear mission – to make it simpler for our customers to buy and use IT that’s fit for their purpose.
44590 Guilford Drive
Ashburn, VA 20147
+1 (703) 554-1600
+1 (800) 929-8331
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram