Getting Started in This New World

This emerging model that tears down boundaries between traditionally stove-piped systems is the future, but for execution, it requires a set of steps — not a single action. Just as company goals and priorities vary, no two companies will follow the exact same roadmap to transform IT. Therefore, this section is comprised of two parts:

A description of the six tasks companies must accomplish to leverage their IT assets in a flexible manner.

A real-world example of how one company implemented a hybrid cloud environment to change the dynamics of IT.

Six steps to increase IT flexibility

To leverage IT assets in a flexible manner, companies must accomplish the following tasks:

Seamlessly integrate existing IT with new cloud-based delivery models. Breaking down the stove pipes among existing applications in the data center and cloud-based services requires a sophisticated approach to integration. Moving data across environments requires metadata mapping that ensures that the integration is accomplished in a meaningful manner.

Migrate to flexible IT and business processes. Processes are designed within and across applications throughout the company and across a supply chain. By separating processes from the underlying implementation, companies can gain better leverage of intellectual property. In addition, codifying processes can help management gain better control over how business policy is implemented within systems.

Create a new model that is holistic so that the underlying services are accurate when used in a new context. This model must be well-controlled, secured, and predictable. One of the biggest problems with traditional IT is that is has been defined based on solving one specific problem rather than on the overall needs of the organization. So, each time a new problem needs to be solved, a new application is written. The new emerging model is to create a set of services that can work together to solve many problems. These services are architected based on the common services — underlying middleware, business process, and data consistent across the company. These services set the pattern for reuse. Organizations are leveraging the models created with a service oriented architecture (SOA) to achieve these goals. SOA patterns provide the best practices that are at the heart of well-defined cloud environments.

Create a lifecycle approach where there is a predictable method of changing elements of business services to support business change. Establishing a more flexible method of managing workloads in different cloud models and on-premises or data center–based delivery models requires that organizations look at the lifecycle of the elements within each environment.

Ensure manageability across the hybrid environment. New delivery models don’t stop at the data center; they incorporate cloud models and delivery models from partners and suppliers. All these services are now the new definition of what we think the role of the IT organization is becoming. Therefore, all these services must be managed as though the composite were a single integrated environment, rather than a highly distributed one.

Provide a predictable level of service across all traditional and emerging cloud models. After an organization takes a hybrid approach based on a combination of different delivery models, creating a consistent level of service is much easier. The new architected environment can now be combined with a service management approach that can be tuned to deliver the level of services customers expect. Not all services will require the same service level. However, in this new service level approach, management can determine which aspects of the environment need more sophisticated levels of predictability, security, and reliability.

Changing the dynamics of IT: One company’s experience

The IT organization of a large multi-national company recognized that inefficiencies in its IT processes were inhibiting its ability to support the business. New and innovative business opportunities were slow to get off the ground because of complicated and lengthy IT approval processes. In addition, the process to requisition new infrastructure would often take so long that teams would purchase more than the project required. As demand for new product offerings grew, the supporting applications were moved from smaller machines to larger ones. Likewise, applications which didn’t take off as anticipated, sat on underutilized large machines. Given the size of the organization this process was repeated again and again, leaving a proliferation of underutilized and ultimately wasted resources throughout the company. Additionally, the company was not able to effectively manage these distributed systems.

The company had two top priorities:

Providing development teams with fast and efficient access to compute resources

Decreasing the cost of maintaining and delivering infrastructure

At the same time, the organization needed to maintain tight control over customer data to comply with industry regulations. The implementation of a hybrid cloud environment allowed the company to achieve its goals and increase speed to market. Therefore, a private cloud was created to manage applications that touched customer data, and SaaS applications were adopted for human resources, marketing, and customer relations. The data center was streamlined by eliminating unnecessary and redundant applications and now houses the organization’s ERP and back-office applications.

The company’s private cloud implementation helps to increase standardization and automation while reducing costs. Their implementation strategy includes these four key best practice components:

Implement a self-provisioning portal: Users at the company no longer need to wait weeks or months to requisition compute power, but can instead be up and running in minutes. Teams are able to do this through a self-provisioning portal that gives employees with the correct credentials access to pre-approved virtual machines.

What advantage does this give to the organization? Ultimately, the ability to pilot new ideas without the need to make large infrastructure investments. Previously, if teams had a good idea, it could take months just to get to a testing phase. Business teams can now offer new products and services to the market sooner and with more effective market response.

Establish a code library: The company created large code libraries including core functionality, common user interfaces, and web services. Without regular attention, registries treated as standalone code repositories can quickly become out of sync. As a result, people stop believing in them and therefore stop using them. To avoid this, a service registry can be built in so that it is automatically in sync.

The code library greatly reduces the time that an architect must spend on coding an application because most of the basic coding has already been completed. Additionally, testing and error-checking efficiency is increased because the majority of the code is standardized and has already been checked. The standardization of application architecture has not only increased speed, but has also driven down the costs of production and increased employee productivity. Overall, the code library provides the organization with greater overall quality by allowing architects to reuse trusted and reliable code.

Increase use of open source software: Open source software allowed the company to realize significant cost savings. They were able to slash many of the service fees that they were previously paying for applications.

Caution must be exercised, however, when adding open source software. Once implemented, it becomes an integral part of the hybrid cloud environment.

Increase commoditization of compute resources: Creating a hybrid cloud allowed the company to tie together hardware infrastructure that can then function as a single large machine. Large, consistent, highly optimized servers can handle larger workloads and therefore take the place of many smaller machines. They require lower heating, cooling, and power consumption, and they can easily be swapped out as more efficient and faster machines become available.