There are different types of data centers. There are data centers that require minimal or even zero capital. There are those that give you full control over your IT infrastructure. There are even data centers that are so small, it’s a wonder they’re still classified as data centers. It only goes to show how much the technology behind data centers have evolved.
Since data centers are integral to nearly every modern, IT-dependent business, those who lead these businesses must be familiar with the different types of data centers. It will help you align your business goals with the right IT infrastructure.
In this post, we’re going to review what a data center is and then proceed to discuss the different types of data centers. Let’s begin.
What is a data center?
A data center is a facility where you can house IT infrastructure, including IT hardware like servers, network devices, storage devices, etc. Of course, that already includes the operating systems, applications, and data stored in those servers and devices. Deploying your IT infrastructure in a data center is much better than just keeping it in the same building that houses your offices and other business operations. To help you understand why, let me explain what a data center does.
What Does a Data Center Do?
A data center will typically have dedicated space, power, cooling, and connectivity for your IT infrastructure. Data centers also provide redundant components for these amenities for those organizations who require high degrees of uptime. Moreover, they employ dedicated security, IT, and facility staff to manage and maintain the entire premises. These attributes enable a data center to deliver significantly higher levels of security, reliability, and connectivity than a regular building.
Now that you know what a data center is and what it does, let’s move on and discuss the different types of data centers.
6 Types of Data centers
In this section, I’m going to give you the lowdown on 6 different types of data centers. I’m going to explain what they are, their major strength(s), and why you might want to deploy your IT infrastructure on them.
- Colocation data center
Arguably the most widely used type of data center, the colocation data center or colo data center is a multi-tenant facility. That means it serves multiple tenants or clients. When you deploy your IT infrastructure in a colo, you’ll be sharing the facility with other clients.
A colocation data center has one major advantage. It’s more affordable than most types of data centers from a CAPEX standpoint. At the same time, it lets you retain a significant amount of control. In a colo, you lease space, power, cooling, etc., from your data center operator but you still have control over your IT hardware and software.
A colocation data center is normally suitable for you if you’re a small business. That said, even if you run larger businesses but want to save on capital costs and don’t mind sharing a facility with other organizations, a colo can be a good option.
- Enterprise data center
An enterprise data center is a facility built, privately owned, and exclusively used by one organization. That would be you, for example. Naturally, to build your own data center, you’ll need a massive amount of capital. A data center’s powered shell alone can cost anywhere around $400M-$800M. A powered shell is the physical exterior plus power and cooling. You’ll also have to assemble a sizable team of management, IT, and security personnel.
Being privately owned and operated, the biggest strength of an enterprise data center is that it gives you complete control over your entire IT infrastructure. That includes your applications and data, all the way down to your hardware and even the facility itself.
Of all the types of data centers outlined here, this is the type you’ll want to choose if you’re a large enterprise and want complete exclusivity in a data center. It’s not going to be cheap. But if you’ve got the financial and human capital for it, building and operating an enterprise data center can be suitable for your business.
- Cloud data center
A cloud data center is a cloud-based service that gives you access to a virtualized IT infrastructure. If you’re familiar with cloud computing, it’s essentially Infrastructure as a Service (IaaS). Among the types of data centers in this list, the cloud data center is the only one where you don’t have to manage the physical aspects of your IT infrastructure. I’m referring to physical servers, routers, networking, etc.
Being a cloud-based service, the biggest advantage of a cloud data center is that it has all the characteristics of cloud computing. Meaning, it’s highly scalable, requires practically zero CAPEX, provides flexible payments, etc.
This type of data center is the ideal choice if you want an option that can give you even greater CAPEX savings than a colo. The downside is that you won’t have full control over the facility and the physical IT infrastructure.
- Edge Data Center
One of the newer types of data centers, an edge data center is a relatively smaller-sized colocation facility purposely built as close to its target end users as possible. Instead of building one huge data center, edge data center operators build many of these smaller facilities in close proximity to their target clients.
This architecture is meant to reduce latency, which improves application response times and the overall user experience (UX). Needless to say, the biggest advantage of an edge data center is speed.
From a network and UX standpoint, an edge data center is faster than traditional, monolithic data centers. So if you’re in the gaming or content delivery business or simply need fast connectivity from your data center-hosted applications, an edge data center would be a good option.
- Micro Data Center
A micro data center is essentially a subclass of the edge data center. These types of data centers are substantially smaller than the average edge data center. Some micro data centers can be as small as a room or even a locker.
Being an edge data center, a micro data center’s strength lies in its ability to reduce latency. This modular design deviates from our definition of a data center above, as it’s no longer a facility. In fact, it’s so small, you would typically house these types of data centers in your facility. I’m not sure why it’s still called a data center. But it exists, so I’m including it here.
You would typically use a micro data center if your operations are heavily dependent Internet of Things (IoT) devices. These devices work best in environments with very minimal latency.
- Hyperscale Data Center
Hyperscale data centers are sprawling facilities that house several thousands of servers in a single site. They’re each equipped with massive capacities for storage, high-speed networks, and computing power. There’s really no clear distinction between hyperscale and colo or enterprise data centers. Hyperscale data centers are simply understood to be significantly more powerful and scalable than the other two.
The main strength of these types of data centers is their ability to support projects that not only demand massive computing resources, but can also scale rapidly. Companies like Amazon, Google, and Facebook all use hyperscale data centers.
Hyperscale data centers aren’t for everyone. I don’t have the exact price range, but the size of the companies that use them should give you a hint. If you’re running big data projects or simply need computing power on a massive and ever-growing scale, hyperscale must be for you.
Another way of classifying data centers is through data center tier ratings, a rating system for data center performance developed by the Uptime Institute. Let’s briefly go over those tier ratings now.
Data Center Tier Rating Breakdown: Tier 1, 2, 3, 4
In this rating system, Tier 1 has the lowest performance capabilities and Tier 4 has the highest performance capabilities. Of course, higher tiers equate to higher costs. According to the Uptime Institute, the tiers are progressive. Meaning, each higher tier includes the capabilities of lower tiers.
In the table below, you’ll often encounter the terms ‘redundant’ or ‘redundancy’. Redundancy simply refers to the presence of backup equipment or system that can take take over should the main equipment or system fail. The more backup equipment/systems there are, the greater the redundancy. Greater redundancy amounts to higher uptime or availability.
| Tier | Brief description | Advantage over lower tier | Is this suitable for you? |
| Tier 1 | Minimal to zero redundancy for power and cooling components. Facility needs to be shut down if components need to be removed for maintenance/repairs. Prevents minor disruptions due to human error, but not unexpected major outages. | N/A | Yes, if you’re a small to medium-sized business (SMB) that doesn’t run mission-critical systems |
| Tier 2 | Provides redundant capacity components for power and cooling. Unexpected major outages can still impact the system | Components can be removed for maintenance/repairs without shutting down the facility. | Yes, if you’re a SMBs that runs mission-critical systems |
| Tier 3 | Greater redundancy than Tier 2. | Unexpected major outages have lesser impact than in Tier 1 and Tier 2. | Yes, if you’re a large business with mission-critical systems |
| Tie 4 | Adds fault tolerance to Tier 3. | Theoretically, when any equipment fails, IT operations can continue as normal. | Yes, if you’re a government agency, hospital, or any organization where downtime can put lives or the general public at risk. |
Final Thoughts
In choosing the right data center for your organization, it’s best to be familiar with the different types of data centers. It will enable you to take into account the major advantages and limitations of each type, and help you make an informed decision.
In this post, we talked about 6 different types of data centers, including:
- Colocation data centers
- Enterprise data centers
- Cloud data centers
- Edge data centers
- Micro data centers
- Hyperscale data center
We pointed out which data center can provide the best CAPEX savings, the greatest control, the fastest response times, the highest scalability, and so on. We also briefly covered data center tier ratings 1,2,3, and 4, and offered information as well as insights around those tiers.
Should you have any additional questions, I encourage you to check out the FAQ and Reference sections below.
FAQ
What do data center operators use for cooling?
The most common cooling systems used in data centers are the Computer Room Air Conditioner (CRAC) and the Computer Room Air Handler (CRAH). The main difference between the two is that the CRAC works almost like your home airconditioning system, whereas the CRAH requires under-floor plumbing.
How to get back in operation if we suffer a major outage?
Some major data center outages, such as those caused by destructive natural disasters, can last long enough to cost your business significant losses. One way to quickly get back in operation is by instituting a disaster recovery plan (DRP) before any outage happens. A DRP will ensure your organization knows how to respond in the event a disaster puts your business out of commission.
Is there a way to detect hackers in my data center network?
Malicious activity like those carried out by hackers can be detected by using intrusion detection systems and intrusion prevention systems. These cybersecurity solutions are equipped with signature-based or anomaly-based intrusion detection methods that identify malicious activity.