|
During the 1980s, IT was considered the panacea for achieving competitive advantage, and it was therefore suitably engulfed in funding. Today, IT is at the very heart of the business, and a company's survival may depend upon the reliability and flexibility of its IT architecture. Ironically, while IT continuity may be critical to the survival of the business and although inadequate IT deployment carries considerable business liability, there has been a parallel shift to reduce IT budgets. The pressure is on to do more with less?to provide more formal service-level agreements (SLAs) with fewer dollars, fewer people, and less infrastructure. With pressure to deliver a cost-effective service that is flexible enough to dynamically adapt to changing business requirements, perhaps the following concerns have crossed your mind:
1. Can I be certain that I have sufficient knowledge of the business direction to predict what the IT requirements will be in 3?5 years?
2. In order to retain some level of flexibility, we constantly juggle the overprovision of IT resources with the need to reduce overall IT costs. A flexible IT infrastructure that is able to adapt quickly seems at odds with our costreduction initiatives.
3. We're looking at implementing SLAs, but can we deliver? Some automated response to an at-risk SLA prior to failure would inspire confidence.
4. Reduced cost is a goal?but not at the expense of availability. Our mission-critical applications require high availability and a disaster-tolerant infrastructure, which inherently demand a costly overprovisioning of servers.
5. We have unused mission-critical server capacity, but we still struggle to meet peak workload demands.
6. Software licensing takes up a considerable part of our IT budget. Ways to reduce these costs would be of great interest. 7. If we could reduce upfront hardware costs and still retain the ability to access additional resources when required (to avoid failing to meet an SLA, as a quick hardware replacement, to fulfill a transient demand, or to evaluate performance), that would be ideal.
8. We juggle resources to create temporary test and development environments. These may be UNIX, Linux, or Windows. Being able to deploy the best operating system for a particular purpose, independent of the underlying hardware each time, would be useful and cost-effective.
9. In theory, the promises of increased flexibility and cost containment associated with IT virtualization are attractive, but:
- Flexibility at the expense of performance is not acceptable.
- Security must not be compromised.
- Wouldn't management be more complex?
- We'd like to gradually introduce functionality into an existing environment.
At HP we have faced these challenges?made even more complex by the acquisitions and mergers now commonplace in business. Consolidation presented HP with the opportunity to adopt a virtualized IT infrastructure?and, in doing so, improve our own ROI. It also enabled us to better understand our IT needs? and to prioritize enhancements to an already mature portfolio of virtualization solutions. HPshopping is HP's e-business, serving home and homeoffice users throughout North America. The challenge was to replace 35 legacy systems, reduce costs, and improve service levels. Adopting an HP-UX virtualization solution resulted in a 50% reduction in IT costs and, at the same time, allowed us to double in size while ranking best in service.
So what is virtualization? Conventional IT architecture assumes a one-to-one mapping between a server and an application: "This is our database server. Over there is our data warehouse, and that one is our CRM server." Each environment is sized for a peak load that may be a single end-of-month run. Not surprisingly, IT studies have shown that the average server utilization rate is an incredibly low 20?35%. Wouldn't it be an interesting approach to break down the physical barriers, pool the resources, and share any excess capacity between applications as they need it? That is exactly what HP does with virtualization. Virtualization is the creation of a pool of physical resources (servers, storage, and network) such that the physical infrastructure is transparent to the end user and the boundaries are flexible enough to shift automatically to meet the end user's changing demands.
|
