Industry surveys have shown the average enterprise spends up to 80 percent of its budget on maintaining existing applications and infrastructure (September 17, 2003). These costs are continuing to grow. The industry is aware of this challenge, and is working to deliver solutions that automate many critical and time-consuming tasks. Yet every IT organization must also address the major sources of cost and complexity in their unique environment. By optimizing infrastructure and operations, you can spend less on maintaining the status quo, and more on new projects and initiatives that deliver direct value to your business.
This article identifies six key strategies for IT optimization. Together, they can put your organization on a path toward continual optimization that delivers better and more measurable business value and lower total costs.
Another critical step is to establish service delivery as a fundamental IT metric (Figure 2). This establishes a quantitative link between key business processes, the IT services they depend on, and the underlying infrastructure of platforms, applications, networks, etc. It helps to bridge the traditional communications gap between IT and business decision-makers, providing a common framework for understanding IT expenditures and the business value they deliver.
The first step is to take stock of your current environment. How many different databases, operating systems and backup and recovery solutions are actually deployed? How many custom applications, management tools, drivers and interfaces? What development tools are used, and how often are design elements reused to accelerate development and increase consistency?
The answers to such questions provide the foundation for moving toward a simpler and more standardized environment, with a limited number of design patterns and consistent architectures, configurations, components and services.
It is not within the scope of this article to discuss individual platform standards. However, at a minimum, consider each of the following when purchasing new platforms and applications: IPMI (Intelligent Platform Management Interface) for server management; DMI (Desktop Management Interface) and ASF (Alert Standard Format) for client management; PXE (Pre-boot Execution Environment), EFI (Extensible Firmware Interface) and ACPI (Advanced Configuration and Power Interface) for all platforms; and WBEM (Web-Based Enterprise Management), which includes CIM (Common Information Model), for enterprise management.
You should also track developing standards, particularly those that will emerge from the Server Management Working Group of the DMTF (Distributed Management Task Force). The work of this group can be expected to provide an important breakthrough for enterprise infrastructure management, greatly improving cross-vendor interoperability for core management functions.
In many cases, these applications are best deployed following, or in conjunction with, projects designed to consolidate infrastructure and/or operations. The combination of platform consolidation and management automation will often deliver better value than either approach independently, providing faster ROI and better project justification.
When planning such upgrades, there are many options to choose from. Industry-standard solutions now include single platforms with up to 128 processors, enterprise blade frames, and clusters with over a 1,000 nodes. Many of these systems also support advanced capabilities that used to be found only in proprietary mainframes, including workload management and virtualization tools that enable dynamic control of platform resources.
In applying the six recommendations outlined above, it is important to consider every project not as an isolated undertaking, but as an integral part of your larger business and IT environment. Invest the time and resources up front to understand business, service delivery and operational needs, and use that information to create detailed engineering requirements. Involve all affected parties at the inception of each project, including business units, financial experts, and IT development, infrastructure, and support staff.
This comprehensive approach can be daunting at first, since it is slower and more costly than organizing a bare bones business and technical team to develop the solution in isolation. Yet the advantages can be immense. On a project-by-project basis, you can expect better buy-in from affected parties, smoother implementation, and a far better chance of achieving real and measurable business value over the life of the solution.
On a larger scale, this strategy can help you move incrementally and cost-effectively toward a simpler and more consistent IT environment based on industry-standard building blocks. As better platforms, applications and management solutions continue to emerge, you will be better positioned to chose and integrate those that deliver the best total value for your business.
Allyson Klein is Initiative Marketing Manager, Enterprise Platform Group, at Intel Corporation.