Look to the Cloud

If you want to see where computing is taking place, look to the cloud. Cloud computing is now the fastest-growing form of computing. According to Cisco Systems, 94 percent of all computing workloads will run in some form of cloud environment by 2021. This includes both public and private cloud plat­forms. Dedicated servers will be a distinct minority.

Cloud computing has become an affordable and sensible option for companies of all sizes, ranging from tiny Internet startups to established companies like Netflix and FedEx. For example, Amazon Web Services (AWS) provides subscribing companies with flexible computing power and data storage as well as data management, messaging, payment, and other services that can be used together or individually, as the business requires. Anyone with an Internet connection and a little bit of money can harness the same computing systems that Amazon itself uses to run its retail business. If customers provide specifica­tions on the amount of server space, bandwidth, stor­age, and any other services they require, AWS can automatically allocate those resources. You don’t pay a monthly or yearly fee to use Amazon’s computing resources—instead, you pay for exactly what you use. Economies of scale keep costs astonishingly low, and AWS has been able to keep reducing prices. To remain competitive, other cloud computing vendors have had to follow suit.

Cloud computing also appeals to many businesses because the cloud services provider will handle all of the maintenance and upkeep of their IT infra­structures, allowing these businesses to spend more time on higher-value work. Start-up companies and smaller companies are finding that they no longer need to build their own data center. With cloud infra­structures like Amazon’s readily available, they have access to technical capability that was formerly avail­able to only much larger businesses. Hi-Media is the Internet publisher of the Fotolog photo blogging web­site. Hi-Media rebuilt the site and moved it to AWS where it can easily scale computing capacity to meet the demands of Fotolog’s 32 million global users who have collectively posted 1 billion photos and 10 bil­lion comments.

Although cloud computing has been touted as a cheap and more flexible alternative to buying and owning information technology, this isn’t always the case. For large companies, paying a public cloud pro­vider a monthly service fee for 10,000 or more em­ployees may actually be more expensive than hav­ing the company maintain its own IT infrastructure and staff. Companies also worry about unexpected “runaway costs” from using a pay-per-use model. Integrating cloud services with existing IT infrastruc­tures, errors, mismanagement, or unusually high volumes of web traffic will run up the bill for cloud service users.

A major barrier to widespread cloud adoption has been concerns about cloud reliability and security. Problems with Amazon Web Services’ Direct Connect service took down several large customers on the morning of March 2, 2018, including enterprise software tool provider Atalassian, Capital One, and Amazon’s own Alexa personal assistant. (AWS Direct Connect is used by hybrid cloud customers to set up a secure connection between AWS infrastructure and the customer’s on-premises infrastructure.) Amazon’s S3 cloud storage service experienced a four-hour outage February 28, 2017, shutting down thousands of websites across the Internet. There were also significant Amazon cloud outages in the preceding five years. As cloud computing continues to mature and the major cloud infrastructure providers gain more experience, cloud service and reliability have steadily improved. Experts recommend that compa­nies for whom an outage would be a major risk con­sider using another computing service as a backup.

In February 2016 Netflix completed a decade-long project to shut down its own data centers and use Amazon’s cloud exclusively to run its business. Management liked not having to guess months before­hand what the firm’s hardware, storage, and network­ing needs would be. AWS would provide whatever Netflix needed at the moment. Netflix also maintains a content-delivery network through Internet service providers and other third parties to speed up the de­livery of movies and web traffic between Netflix and its customers. Netflix competes with Amazon in the video-streaming business, and it wanted to retain con­trol of its own content delivery network.

Dropbox, on the other hand, did the opposite.

The online file hosting company saved nearly $75 million in infrastructure costs over two years following a cloud data migration off AWS. Dropbox had been an early AWS success story, but it had never run all of its systems on AWS. Dropbox had originally split its architecture to host metadata that provides information about other data in private data centers and to host file content on the AWS Simple Storage Service (S3). Dropbox subsequently built systems better suited to its needs, which so far has produced big savings following its cloud data migration off AWS. However, that transition was costly. The company spent more than $53 million for custom architectures in three colocation facili­ties to accommodate exabytes of storage. Dropbox stores the remaining 10 percent of user data on AWS, in part to localize data in the United States and Europe, and it uses Amazon’s public cloud to help deliver its services. Experts believe that Dropbox’s experience with AWS is not representative of most companies. Dropbox’s strategy to build one of the largest data stores in the world depended on owning its computing resources.

Many large companies are moving more of their computing to the cloud but are unable to migrate completely. Legacy systems are the most difficult to switch over. Most midsized and large companies will gravitate toward a hybrid approach. The top cloud providers themselves—Amazon, Google, Microsoft, and IBM—use their own public cloud services for some purposes, but they continue to keep certain functions on private servers. Worries about reliabil­ity, security, and risks of change have made it diffi­cult for them to move critical computing tasks to the public cloud.

Honda UK implemented the hybrid cloud model to enable its IT infrastructure to handle sudden spikes in usage of its websites. The company had experienced sudden web server crashes due to band­width limitations. Honda UK had initially moved to a private cloud model, which was used during the launch of the Accord Thurer model to handle heavy user demand for its website. Honda UK then started using the public cloud during the launch of the Honda CR-Z. Honda UK had to pay for the cloud service only when the company used it. The pay-as- you-go model helped keep costs in check while en­suring optimum scalability.

Source: Laudon Kenneth C., Laudon Jane Price (2020), Management Information Systems: Managing the Digital Firm, Pearson; 16^th edition.