Data Center As A Service (DCaaS), Colocation Offer Data Center Flexibility, Expandability

Data Center As A Service (DCaaS), Colocation Offer Data Center Flexibility, Expandability

First part of a four-part article examining the potential benefits of DCaaS and colocation.

By Paul Schlattman  
OTHER PARTS OF THIS ARTICLEPt. 1: This PagePt. 2: Tips for FMs Who Run Colocation and DCaaS Data CentersPt. 3: QTS Chicago: Colocation Data Center Features Low Latency, High Reliablility, and Redundant InfrastructuresPt. 4: A Brief History Of Colocation

According to Cisco, the world now processes 2.5 billion gigabytes of data daily. A large piece of this demand is due to the introduction of Internet Protocol (IP) over everything, generally described in the industry as the Internet of Things (IoT). Today IP addresses are being attached to practically every electronic device manufactured. Cameras, lights, doorbells, computers, and appliances — and of course intelligent building systems — all contain IP devices to manage or view electronics over the cloud. Cisco anticipates that the economic value of IoT in 2022 will exceed $14.4 trillion.

This great demand for IoT requires data centers to be able to grow as needed and be supported by cloud applications. As more enterprise companies continue to outsource their processing to colocation facilities, the capabilities of colocation data centers will need to be flexible enough to accommodate myriad new technologies; in many cases the colocation and wholesale data centers of the past are not designed to be this flexible in supporting such applications. New data centers that are now being designed for IoT are labeled data center as a service or DCaaS. 

A DCaaS facility — where state-of-the-art equipment, space, and bandwidth are available for rent — can provide a great opportunity for companies large and small to take advantages of DCaaS services to manage future changes in technology. When subscribing to a colocation (co-lo) company that offers DCaaS, tenants can request rack installations and other services to quickly support new IoT applications. DCaaS has become the third-generation data center era. (The first two were standard colocation and wholesale.)

The emergence of DCaaS means that organizations should evaluate their enterprise data centers to see if a co-lo/DCaaS is a good option. And the facility manager will have a new role in the co-lo/DCaaS.

Understanding third generation co-los
Today’s third generation co-lo data center looks remarkably different than the first generation co-lo of the dot com era. For one thing, the third generation co-lo data center is built to support hyperscale compute (which is required by very large data center users like Microsoft or Apple), including offering cloud computing services as well as scalability of both processing and storage. Depending on the market, industry and Kw infrastructure needed, today’s third generation co-lo is ripe to replace the legacy, enterprise data center.

The role of the facility manager in determining whether to move compute capability from an enterprise data center to DCaaS will require the creation of a formula for total cost of ownership (TCO). The formula should consider the following needs that the company may have now or may require down the road. The in-house facility manager responsible for the enterprise data center should help the organization address these issues by evaluating the existing facility infrastructure of the enterprise data center and by analyzing the facility infrastructure of the co-lo/DCaaS provider to ensure that it can meet the needs of the organization. Consider the following:

• Mobile applications. If the organization is already engaging in or moving toward use of mobile applications, investigate what processing needs there are. Does the enterprise data center facility have these capabilities? If not, what upgrades would be required? Can the organization sustain them cost effectively?

• Driven by devices. For manufacturing companies that are constantly rolling out new applications associated with their products, consider future device-driven IoT processing infrastructure requirements.  

• Cloud computing. Moving some of the processing power to the cloud not only reduces the load on the enterprise data center but also improves efficiency and may reduce costs. Other benefits include flexibility, disaster recovery, and automatic software updates.

• Latency requirements. Identify the latency requirements of the new application. Can the enterprise data center provide faster speeds or would a colo/DCaaS speed up response/delivery to customers?

Some organizations will find a happy medium by utilizing an existing enterprise data center along with DCaaS and cloud computing. For example, take a department store with 2,400 applications that need to be supported. The organization could allocate 120 or even 300 applications to the cloud, while keeping more than 2,000 in-house at the enterprise or first- or second-generation co-lo data center. Now, the total operating load has been trimmed down and therefore, the TCO of the in-house applications is less.

In the case of the typical enterprise company, the formula for TCO would look like this: TCO = enterprise data center operational costs + cloud computing costs + co-lo/DCaaS costs.

Contact FacilitiesNet Editorial Staff »

  posted on 2/10/2017   Article Use Policy

Related Topics: