Public Cloud Drives Enterprise Data Center Consolidation

New breed of enterprise software and cloud capabilities force companies to rethink infrastructure strategy

Yevgeniy Sverdlik

October 8, 2015

5 Min Read
Jeff Bezos Amazon
Amazon CEO Jeff Bezos at a press conference in Santa Monica, California. (Photo by David McNew/Getty Images)Getty Images

Amazon Web Services, which just a few years ago was a provider of rentable virtual servers thought to be a good fit for startups and test and development workloads only, is a different beast today, working hard – and in many cases successfully – to become the cloud enterprises migrate their corporate data centers to.

The need for new kinds of applications and availability of cloud services is forcing companies to take a hard look at their IT infrastructure and the cost of operating it, and many of them find that it doesn’t make business sense to keep managing on-prem data centers, or at least that it doesn’t make sense to manage the amount of on-prem data centers they manage, and that they can get away with managing fewer.

AWS isn’t the only public cloud provider fighting for the hearts and minds of enterprise IT of course. Its rivals like Microsoft and IBM are making substantial inroads in the enterprise cloud market and have built formidable alternatives to the public cloud giant.

But the giant is making a lot of money in the market and claims its revenue is growing faster than anyone else’s. In this year’s second quarter, the revenue run rate of Amazon’s cloud business was $7.3 billion – about 80 percent up year over year, according to Andy Jassy, senior VP for AWS, who delivered the opening keynote at the AWS re:Invent conference in Las Vegas Wednesday.

“We are seeing a very significant uptake in public cloud use across enterprises,” Richard Villars, VP of data center and cloud market research at IDC, said. In surveys and interviews, IDC sees a much broader commitment among enterprises to public cloud as a platform for new applications and as a driver for data center consolidation plans. The latter is a more recent trend which has become visible over the past eight months or so, he said.

Companies IDC is talking to have plans to consolidate corporate data centers and move workloads to AWS, Microsoft Azure, IBM SoftLayer, and in some cases Google Cloud Platform. Public cloud isn’t the only option of course – companies are also increasingly looking at managed services and hosting – but the primary drivers are similar: “People just want to get out of the data center business,” Villars said.

Market Realities Demand Infrastructure Rethink

C-level technology execs from General Electric and Capital One, two companies among some of the largest enterprises, said they are implementing data center consolidation plans that consist of shutting down most of their data centers and moving application workloads into AWS. The two execs participated in Wednesday’s keynote as speakers.

Rob Alexander, CIO at Capital One, said the bank is going from operating eight data centers in 2014 to five in 2016, and then to three in 2018. Moving most data and applications to the cloud enables Capital One to roll out new software faster, have elastic capacity that can expand during high-demand periods, such as Cyber Monday, and contract when demand is lower, he said.

Capital One has realized it can now deploy some of its most critical production platforms on AWS, according to Alexander.

Building and deploying new applications and delivering them in new ways are a big part of the reasoning behind moving to cloud infrastructure. For banks, delivering software products on mobile devices is crucial. “Mobile is moving fast,” Alexander said. “Mobile has become the preferred channel for our customers.” Capital One’s latest mobile banking application runs on AWS.

Banking is one of the industries where the ability to build and deploy software quickly has become a primary way to stay competitive. “We have to be great at building software and data products if we’re going to win at where banking is going,” Alexander said.

Connected Devices Need Distributed Infrastructure

Enterprises are spending more and more on outward-facing applications that deliver content and services to customers on mobile devices in new geographies, and “if your data center is sitting in your headquarters, it’s probably in the wrong place,” Villars said. The distributed nature of public cloud infrastructure allows companies to deliver those services from data centers closer to the users.

GE, while a very different business from Capital One, is also finding that public cloud is a better infrastructure option than running everything internally. In some ways, its needs are similar to Capital One’s: it also has to send and receive data from end-point devices it hasn’t communicated with in the past. Except, for GE those end points are things like wind turbines, aircraft engines, or machines at manufacturing plants, Jim Fowler, the industrial giant’s CIO said while on stage at re:Invent.

GE’s ongoing data center consolidation program is expected to take the company from 34 data centers to four, which will hold only the most secret and valuable data it has, Fowler said. The rest – thousands of workloads – will go to AWS over the next three years.

Besides supporting internal IT, GE recently became a cloud service provider itself, albeit its cloud services are of a different breed than those AWS provides. Its Predix cloud is a Platform-as-a-Service offering that interconnects devices in manufacturing facilities and helps optimize operations by leveraging data analytics. GE calls it the “industrial internet.”

Cloud Feature Set Grows to Speed up Transition

Public cloud is becoming more and more attractive for enterprises, and cloud providers and the myriad of companies that provide services around the big public cloud platforms continue rolling out features that make it increasingly so. A number of the new service announcements Jassy made during Wednesday’s keynote were aimed at helping enterprises move applications from on-prem facilities to the cloud or start using cloud-native services.

Amazon Kinesis Firehose, for example, is a service that allows companies to quickly load streaming data into AWS for analytics. The new AWS Database Migration service enables them to easily migrate production databases to AWS with minimal downtime. AWS Config Rules allows enterprises to set up compliance rules for resource configuration and get alerts when the rules aren’t being met. Amazon Inspector is an automated security assessment service that identifies security or compliance issues on AWS, assessing network, VMs, OS, and app configuration.

With the new Amazon Snowball service, AWS will ship any number of ruggedized, temper-proof 50TB storage servers directly to your data center and then back to one of its cloud data centers after you’ve loaded them with your enterprise data to be uploaded into AWS. The service is meant to speed up data migration from on-prem data centers to AWS, which at that scale can take several months if done over a network.

Subscribe to the Data Center Knowledge Newsletter
Get analysis and expert insight on the latest in data center business and technology delivered to your inbox daily.

You May Also Like