A lire sur: http://searchcloudprovider.techtarget.com/tip/Choosing-cloud-infrastructure-for-business-and-technology-flexibility
Editor's Note: In part one of our guide to enabling cloud services, we assess the business and technology issues that should influence providers' cloud infrastructure choices for deploying cloud services that can change over time as business needs demand. From resource allocation to service management to billing, cloud providers need to find the business and technology balance for the cloud service model that appeals to the enterprise buyer yet enables providers to make a profit.
Cloud computing is one of the hottest topics in information technology because it expands the existing notion that the Internet looks like an actual cloud. It also creates a real cloud business and technology model, which requires that providers make well-informed cloud infrastructure decisions so they can supply elastic computing resources. That way, they're creating new network service opportunities and addressing enterprise business needs for application support and productivity growth.
If the mission of a cloud provider is to offload current enterprise data center activity, the revenue opportunity is merely huge. But if the mission is to create a new service model that distributes features everywhere a user can take a mobile device, the opportunity is staggering. The total global revenue opportunity for cloud computing services is in the hundreds of billions of dollars, and the profit margins for cloud services could be significantly higher than those found in mobile and wireline telecom services.
At its core, cloud computing is based on the assumption that a large pool of resources can be used to run applications at a lower cost and support them more efficiently than a series of application-specific or even company-specific resources. Ten separate servers might support 10 specific applications, but 10 servers in a server farm -- provided they have effective resource allocation and management tools -- could support 20 applications or more. The same servers, because they are centralized, can be supported more efficiently.
Earning a profit while meeting enterprise customer needs
The goal of the cloud is to produce economies of scale in resources and to allow a provider to offer a service that saves the buyer money, but at the same time earns a profit for the seller. Striking this win-win balance is the central mission of deploying cloud infrastructure and platforms, and that mission can be fulfilled in three ways:
- Cloud services have to target prospective customers that will benefit the most from the cloud value proposition;
- The potential customer base -- meaning the business size, type and current technology commitments -- shapes the benefits that buyers need and understand. That means services will have a specific opportunity target, which will demand a specific technology focus.
- The cloud market will mature like all markets do, and it is likely that providers will have to continually broaden their service range to broaden customer and revenue opportunities.
For cloud providers, the challenge is to make these assumptions about cloud efficiency real, and doing that isn’t simply a matter of building a data center and linking it to a delivery network.
Cloud infrastructure must offer both business and technology benefits
Cloud-specific technology components are designed for the kind of evolution that originates with certain services and then broadens them. While every company in the space doesn’t necessarily offer the full range of solutions, every cloud component company has products that are designed to be integrated into a flexible cloud infrastructure. By selecting a cloud vendor with this kind of flexibility, cloud providers are buying insurance against a predictably dynamic and sometimes disorderly cloud market evolution.
Key components for creating the cloud services business case
The following components of cloud infrastructure provide a specific benefit for making the cloud business case real and helping it evolve:
- Resource allocation. All cloud computing includes a mechanism for resource allocation, management and sharing. A pool of resources is shared among cloud service buyers, and it is critical that these "multi-tenant" resources appear to be dedicated to each buyer. But all of the available resources are efficiently distributed, with at least some dynamic assignment capabilities among cloud users. Cloud resource management is the basis for all cloud services, and it can be based on technologies like virtualization, multithreaded operating systems, service-oriented architecture or a combination of these features.
- Cloud service models. Three basic models of cloud service have emerged: Infrastructure as a Service (IaaS), Platform as a Service (PaaS) and Software as a Service (SaaS). Each of these models cedes some traditional IT infrastructure to the cloud, with SaaS ceding the most and IaaS ceding the least. Since the user must contribute any IT components that are not provided by the cloud, SaaS does the most for the user and displaces the most direct user costs, while IaaS displaces the least.
- Cloud provider profits. Another important element of cloud mission fulfillment is to find a service that lets the provider displace enough cost to increase the profit opportunity but at the same time control provider costs and expand the early market opportunity. Higher-level services like SaaS are more application-specific and therefore have more limited prospect pools. The more that a cloud provider's infrastructure can support all three models, the greater the flexibility it has for early targeting and later service evolution, even though its initial costs may be higher.
- Cloud management and billing. Management and billing are critical to the cloud, and here the most critical issue is multi-tenancy. By definition, the cloud means using shared infrastructure, and the process of sharing means that users can access a pool of resources rather than dedicated resources. Cloud-customer use of shared resources must not compete in an adverse way, and resource management can never be allowed to interfere with the operation of other users. And, finally, customer use of resources must be accounted for in order to manage capacity planning and billing.