“Cloud computing” is one of those “hyped” concepts the telecom industry runs into from time to time that is confusing both for customers and service providers alike.
To make matters worse suppliers often try to repackage existing services by calling them “cloud” services, when, strictly speaking, those services are not “cloud computing” apps or services.
For example, cloud computing implies use of data centers. But data center hosting is not, itself, a cloud service. Data center hosting can involve service level agreements, redundancy, load balancing or guaranteed bandwidth are services but none of those are cloud computing.
Data centers make cloud computing possible, but are not themselves a cloud service.
“Cloud computing” is defined as “Internet-based computing where large groups of remote servers are networked so as to allow sharing of data-processing tasks,centralized data storage, and online access to computer services or resources.”
Wikipedia defines it as “delivery of computing as a service rather than a product, whereby shared resources, software, and information are provided to computers and other devices as a metered service over a network (typically the Internet).”
You can see why the term is confusing. Cloud computing is a marketing term; a delivery model for information technology; a computing architecture; a business model and also the framework for many different types of services.
To make matters worse, many of the components of a “cloud service” resemble older products and apps. “Remote access” to servers, a distributed architecture; computing as a service; data centers; managed services and usage-based billing are building blocks for cloud computing.
Some might say that, although it rarely is a part of the formal definition, cloud computing involves the use of standard browsers as the end-user “client.”
From a customer point of view, only some elements of cloud computing matter, arguably the features, cost and delivery model. By definition, a “shared” approach to network resources means higher efficiency for the provider and therefore lower cost for the end user.
In practice, that means small organizations can afford to use applications that would have been prohibitively priced, if delivered in the old way.
For large organizations, the management advantages also are key. If an app has to be updated, there is no need to fiddle with software on all client devices. The latest version is updated on the servers.
If end users grow or contract, cloud apps simply scale gracefully as well. And storage of data off-site, with backup, can have data integrity advantages as well.
The confusion about “what” cloud computing is, on the part of buyers and sellers, is that it involves several potential changes in “how” users consume computing resources and software, and that many of the changes resemble other processes.
Consider “applications,” “platform” and “infrastructure,” all of which can be delivered as “cloud services.”
Applications, which are seen as the largest cloud computing revenue drivers at the moment, are synonymous with “software as a service,” where users buy access on a subscription basis, instead of on an “owned” basis.
Think of the difference between buying a DVD and renting a move. “Renting a movie” is SaaS.
“Platform as a service” generally is useful to application developers rather than retail end users, as it includes rental of computing resources to create, test and manage an application.
“Infrastructure as a service” is where most “telecom” sales organizations will encounter cloud computing as a product to sell a business buyer. IaaS generally means renting end users computing power and storage, so the notion of “data centers” and “rented computing resources” are key.
But it is the rented computing resources that are “cloud services,” not the use of data centers. Selling data center rack space, power and cooling are growing businesses for telecom companies, but are not “cloud services.” Data centers are an input to create and deliver cloud services, as are the fiber backbone networks that move the data.
Iaas services themselves generally fall into three or four buckets: “public cloud,” “managed public cloud,” “virtual private cloud ” and “private cloud.”
The first two concepts involve end user access to cloud apps on either an “un-managed” basis (typically retail consumer apps such as Google search) or “managed” basis with service level agreements, and as their “public” moniker implies, generally are retail and consumer apps designed for use by the public.
The last two concepts involve something more like traditional enterprise computing. A virtual private cloud simply designates resources to a single customer. It typically is used by an enterprise for “private” network apps internal to the organization.
A “private cloud” typically means an enterprise owns its own servers and uses them to deliver internal data and apps to enterprise employees and partners, not the general public. In some cases, enterprises actually own their own servers but use a remote data center to operate them.
Private cloud is one of the concepts that accounts for confusion. Strictly speaking, it is not cloud computing, since the resources are not rented, but owned. Only the data center facilities are rented.
So what does that mean for telecom service providers and sales forces? It can be very hard to explain precisely what a particular “cloud service” is, or why it is better than a conventional “shrink wrapped” software purchase.
The analogy many will be familiar with is “hosted PBX” services as an alternative to buying a business phone system. That has not proven to be “easy” to explain, either.