Get Stunning ROI from Cloud Computing

The fiscal argument for cloud computing is very strong

Service Management in Action Cloud computing is much more than just a new platform of service delivery. It's also a persuasive response to business-centric questions such as: How can we drive service levels up while driving costs down? How much will we have to invest to make that happen? And how long will it take to see a return on that investment?

Technology executives today are understandably gun-shy of significant expenditures or platform shifts, and cloud computing does imply a modest initial investment in service management and virtualization software. However, according to a 2009 Cloud Computing ROI study conducted by IBM Research, the many benefits of cloud computing amply and quickly justify that investment.

Furthermore, for cost-conscious executives, a pleasant surprise may be in store. While cloud computing cost-savings are often associated primarily with hardware, even more savings will usually stem from reduced labor costs. In fact, according to the IBM study, a remarkable 81 percent of cloud payback is typically achieved in this way.

How is this possible? Primarily, it is because IT services running in a cloud can largely be automated, requiring little or no manual effort to create, deploy, manage and retire—essentially every stage in the service lifecycle. Because the cloud infrastructure is more and more automated to be flexible, scalable and resource-aware, it can create enormous business value while requiring minimal attention from IT. That, in turn, implies not just that mission-critical IT services perform better; it also implies that IT professionals, funds and resources are freed to attend to other, higher-priority tasks like innovative new services to increase customer satisfaction, market share and revenues.

Achieving those compelling results, however, will typically require a service management implementation, in which the cloud infrastructure is managed and governed with a focus not on the underlying technology, but rather the services themselves and how well they're hitting dynamically-changing business targets.

Further, maximizing cloud ROI in the shortest timeframe will require considering the five areas of payback potential IBM has identified (hardware, software, productivity, provisioning and system administration) with respect to the complexities they involve and how those complexities will apply to any organization and its unique context.

Cost savings: Hardware

"Based on customer environment and cost-saving data, it appears that for most organizations, cloud computing delivers a substantial and exceptionally fast return on investment via optimizations in five different areas: hardware, software, automated provisioning, productivity improvements and system administration. " Hardware is perhaps the most obvious example of cloud computing cost reductions. Cloud infrastructures are highly virtualized, running multiple virtual systems per physical host, and thus require fewer hosts. They also deliver superior hardware utilization per host. Instead of many hosts, each of which is utilized to less than 25 percent of its capacity, there are fewer hosts, each of which is utilized closer to 50 percent of capacity. And just as hardware costs are lower, so too are the associated energy costs, both in terms of the power the hosts require and the heat they generate which must be dissipated.

What kind of overall hardware cost-reduction is possible via a cloud? For medium-sized clouds, IBM estimates roughly 62 percent savings in one year; for larger clouds, it will be closer to 50 percent.

Consolidating and virtualizing servers in this way is a great way to begin a cloud deployment project, reduce operational costs, and render the infrastructure generally more agile, dynamic and responsive to unpredictable conditions.

And toward this end, IBM offers many services that are directly applicable, in areas ranging from initial assessment to scope definition to workload evaluation to a comprehensive business case designed to project, quantify and prioritize the many complexities involved, their anticipated ROI and the time required to achieve it.

Further services are available in specific areas of potential optimization such as power/cooling, space utilization, application performance and automation. And, of course, IBM also offers a broad range of leading hardware solutions admirably suited to cloud deployments, such as the integrated IBM CloudBurst offering that combines essential hardware, service management software, storage, networking and setup/integration services to help organizations create a powerful private cloud with minimal time or effort.

Cost savings: Software

Software is the key to a successful, cost-optimized cloud platform. Despite the fact that initial expenditures will be required for new software in order to manage both the virtualized infrastructure and the services it supports, that software will in turn make possible substantial improvements in overall service levels while minimizing the required resources. And those enhanced service levels will, in turn, lead to a rapid and substantial return on the software investment.

Why should new software be necessary? Consider that, among other reasons, traditional management tools will typically have features, and be built on assumptions (such as one-to-one server/application deployment), that no longer apply. A virtualized infrastructure such as a cloud requires management tools that were intended for it, and can capitalize on its strengths for business value.

Organizations looking to create such business value will need to invest in three classes of software for the cloud:

(a) virtualization software itself, which serves as the foundation by which virtual servers are hosted within the cloud, allowing them to run in parallel on any given machine;

(b) service management software, that shifts the focus from the technology to the services supported by the technology, then gives the organization the visibility, control and automation needed to tune those services for highest performance, scalability, efficiency and cost-efficiency;

The IBM service management portfolio collectively comprises a best-in-class, modular set of solutions, built on open standards for seamless interoperability through which organizations can achieve optimized service management in a virtualized cloud infrastructure.

Cost savings: Productivity improvements

Cloud environments also deliver truly dramatic ROI through improvements in productivity. These are made possible thanks to a far more rapid and automatic service creation and management process.

Rather than the traditional, lengthy process of manually requesting a service, and waiting for IT to fulfill all the stages involved in creating that service, users can simply request services directly from the cloud and let the cloud fulfill them. If driven by appropriate software—to provide a user-accessible Web self service portal and catalog of services, and manage those services and their resources subsequently at every stage—clouds can reduce service creation times from weeks or months to less than one business day.

The fiscal outcome of this optimization is impressive, to say the least. For clouds of all sizes, the productivity payback in one year is estimated at a staggering 97 percent. Meanwhile, services will be rolled out far more quickly, governed more comprehensively and the resources they require will be continually put to good business use through automatic allocation.

One powerful solution designed to deliver such remarkable productivity cost-savings is IBM Tivoli Service Automation Manager. This tool accelerates the service request approval process via automated workflows—allowing end users to request and receive new services, that will be managed in a business-optimized manner at every stage, despite not knowing anything about the technical resources involved.

Cost savings: Automated provisioning

Impressive cost-reduction opportunities arise from automated provisioning, used to provision virtual servers with the appropriate and complete software stack (OS, applications, middleware, data, drivers and other elements) they require to drive their services.

The traditional approach to server provisioning—manual—is by comparison far slower and more error-prone, and as a result, much more likely to result in a diminished business outcome because servers may not always have the software they require. That can lead directly to service/application downtime, security shortcomings, compliance failures and many other unwanted results. Virtual servers provisioned automatically, on the other hand, on the basis of script-driven business policies, are provisioned dramatically faster and more consistently. This translates into relatively swift, and high, payback for automated provisioning technology—and a stronger business case for cloud ROI in general.

How much stronger? The larger the cloud, the stronger the case. Typically, the time required for provisioning will fall, through automation, from 40-70 hours to a startling 30 minutes. The more servers in place in the cloud, and the more frequently they must be provisioned, the more optimization automated provisioning will deliver. IBM estimates that for medium-sized clouds, provisioning costs will fall by roughly half; for large clouds, however, they will fall by as much as 90 percent.

Perhaps the best tool available to pursue automated provisioning in a cloud is IBM Tivoli Provisioning Manager, a script-driven solution that allows IT to provision virtual servers based on business policies for an optimized outcome. In addition to its own best-in-class feature set, which includes several forms of automation, this tool also smoothly integrates with Tivoli Service Automation Manager, to create the virtual system(s) required to support a user’s requested service (once that request has received IT approval).

Cost savings: System administration

Finally, the case for cloud ROI is bolstered still further by new possibilities in system administration.

In a cloud environment, physical hosts are fewer, but each may contain many virtual servers. This increases overall management complexity and means new, virtualization-aware tools will be needed to leverage the infrastructure for best business value. Once these are deployed, however, the cost savings are substantial—up to 36 percent savings, for instance, in the case of medium-sized clouds.

IBM offers a broad array of compelling offerings to solve different aspects of the system administration puzzle. One such is IBM Tivoli Application Dependency Discovery Manager, which can automatically discover the chain of dependencies involved in any given application, however it may span the virtual infrastructure, then reflect that chain via a topology map, and reports, for faster time-to-solution in the event of a problem and simplified regulation compliance.

Another: IBM Tivoli Monitoring for Virtual Servers, which delivers key information about virtual server status levels via a centralized portal, empowering IT to preclude many problems from having a business impact at all. And IBM Tivoli Business Service Manager helps by portraying cloud services from a business perspective—showing how cloud services/applications are fulfilling business goals as measured by key performance indicators and depicting the results in graphs, charts, scorecards and other formats.