Costs Of Cloud Computing At Odds With Growing Complexity
Cloud customers are dealing with confusion over cloud computing costs and the complexity of the services they are paying for
A little more than a decade into the public cloud era, the costs of cloud computing are still far from settling into predictable and repeatable patterns.
The reason is why cloud computing became popular in the first place. Users can free themselves from IT CAPEX—capital expenditures—which for even mid-sized businesses can result in millions of dollars of up-front costs for a data center refresh, staff salaries and yearly maintenance fees.
Costs for cloud-as-a-service come out of the OPEX budget—operating expenses—which allows companies to pay only for what they use. That model is very attractive, but things can get tricky, as spikes in demand for some cloud-based applications can get very expensive if users are not prepared. Reserved instances, which are paid for up front, as well as serverless computing, can further complicate matters.
The resulting battle between vendors over the costs of their services can be confusing. On the one hand, costs are going down. Providers like Amazon Web Services regularly lower prices for their services and Larry Ellison has stated that Oracle will not be beat on price by AWS.
Battling FUD and lock-in
But on the other hand, traditional vendors, which still own the lion’s share of workloads on-premises data centers, are walking a fine line between embracing the cloud and protecting their on premises business by stoking fears about the costs to use or move data into or out of the cloud.
For instance, Michael Dell raised a few eyebrows at the Dell EMC World conference when his keynote turned to the subject of the costs of public cloud computing vs. on premise. “Many customers have already told us that public cloud is twice as expensive as on premise,” said the CEO of Dell Technologies, “especially for a predictable portion of their workloads, which in many cases is roughly 90 percent.”
In another example, HPE CEO Meg Whitman recently compared the cloud to the “Hotel California,” referencing the renowned Eagles’ song that says “you can check out any time you like, but you can never leave.”
Some recent research bears this out. At the recent Code Conference, Kleiner Perkins analyst Mary Meeker presented survey data that shows data security and “uncertainty of cost and savings” in the cloud remain major concerns of IT managers, though less so than five years ago. Cloud “lock-in” is a concern for 22 percent of respondents, up from just 7 percent in 2012.
As a result, users are increasingly turning to the growing ecosystem of software and service providers, including CloudHealth, Cloudability and RightScale, whose primary task is to help them understand their cloud bill, regulate usage to avoid surprises and predict future costs.
“We help provide long-term trend data of what [users] are doing across different applications and then we help them project forward what they can expect to be using over time,” said Joe Kinsella, CTO of CloudHealth Technologies. “This allows them to do the computations that help them arrive at a rough approximation of what they will spend for a workload.”
It’s all about the workloads
The myth that needs busting amid the talk around cloud CAPEX vs. OPEX is that cost is not as important as it used to be when it comes to the cloud. The promise of lower costs brought customers to the cloud, but they are staying because of agility and operational efficiencies.
In addition, customers are getting smarter about cloud “all-in” strategies. Many applications, such as CRM, in the case of Salesforce.com, or email or document management, in the case of Microsoft’s Office 365, are tailor-made for the cloud. Other applications, which could be proprietary or constrained by data or compliance requirements, may be better suited to the on premises data center.
Continues on page 2…
