Data Centers: time to regroup and consolidate

In the past decade, IT organizations have undergone a rapid and organic growth in technology, forcing the staffs to continually scramble to meet the ever-increasing demands of the business.

The result is a landscape of multiple data centerslarge and small, scattered across the enterprise, each with a significant population of grossly underutilized physical assets. Some of the assets may be located outside the data centersbranch offices, storage closets, or employees homes. This rampant decentralization has inevitably resulted in fragmentation across those locations.

The resulting inefficiencies have driven up the capital and operating expenses and have made the managing of IT resources even more complicated. To make matters even worse, many data centers are running out of space and are consuming vast amount of electrical power.

Its time to regroup and consolidate your data centers to make them more efficient, agile, manageable, and less costly to operate. Consolidation requires more than just adopting virtualization, cloud, and other related technologies. It involves developing, implementing, managing, and continually improving a strategy that addresses consolidation by doing the following:

Four Steps to Success

Data center consolidation is a major undertaking that should be approached incrementally. Moreover, it is not a one-time activity. A concerted effort at continual improvement is vital in achieving and maintaining optimum utilization of people and technology resources, once consolidation is underway.

#1 Discover Your Assets

Before you can consolidate, you need to understand your data center landscape. Several steps are required to get a full picture of all your assets ie, where they are located, and how they interconnect.

Gather detailed information about your assets: This involves knowing what assets you have in your infrastructure including servers, storage devices, network devices, multi-tiered applications, and middleware

Identify asset locations: You need to know not only the data center in which each asset is located, but also where its located within that data center. This requires mapping the floor layout of each data center, pinpointing the exact location of each hardware asset, and indicating its power consumption.

Define the physical and logical dependencies of the assets: This involves describing their relationship to the business services they support

Combine all of this information into a single, unified data repository: All functional groups in IT should have access to this information

#2 Plan How Your Assets Will Be Used

Once you know what assets you have, where they are located, and how they are being used, you can then determine how to consolidate them to ensure their most effective and efficient use. This requires careful planning. The goal is to determine the best plan based on capacity analysis, application dependencies, and business requirements.

Keep in mind that consolidation provides an invaluable opportunity to rethink how you deliver and configure services.

The planning step involves 5 tasks:

Plan capacity: Begin by determining how much capacity you have and how much of it is being used effectively.

Analyze performance: Its important to maintain or improve performance levels during and after consolidation. Start by analyzing the performance of your assets to establish baselines

Identify consolidation tasks: Capacity planning and performance analysis provides the information you need to perform this third task, which is to identify which hardware assets are candidates for virtualization or transition to the cloud. A proper capacity and performance analysis will indicate how to combine workloads on the virtual systems to use the consolidated capacity during peak times, as well as over the entire day

Develop consolidation scenarios: Once you have identified virtualization/cloud candidates, you are ready to investigate different scenarios for consolidating data centers and for consolidating assets within data centers. Here you investigate merging the capacity of multiple locations into fewer locationsperhaps even a single data center.

Some applications may currently span multiple data centers: With respect to consolidating virtual resources onto physical resources, be sure to investigate various mixes of physical, virtual, and cloud servers

Create an optimal consolidation plan: Compare the possible consolidation scenarios. For example, you can compare the various mixes of physical, virtual, and cloud assets to determine the optimum combination

#3 Implement Your Plan

Implementing your consolidation plan involves 3 levels of effort

Consolidate the infrastructure defined in your plan: This includes deploying the full server stacks for each physical serverbare metal to applications

Virtualize the systems that you have identified: This involves deploying the hypervisor platforms you have chosen. Its important to maintain the flexibility to deploy heterogenous hypervisor platforms so that you can choose the best fit for each virtualized system

Deploy the business services that run in the consolidated environment: By automatically configuring all the layers of the data center in a consistent way and by holding to configuration policies over time, you can reduce downtime, decrease costs, and lower the risk of change. Additionally by automating service delivery, you make it easier to move into cloud computing, IT-as-a-service type of scenario

#4 Manage the Consolidated Environment

Only with effective management you can realize the full benefits of data center consolidation. Be sure to address the following areas

Manage the lifecycle of cloud services from request through retirement: To begin, implement a full lifecycle management of virtual and cloud based systems. This includes management of the building, maintenance, and retirement of those systems.

Ensure continuous compliance: Your consolidated environment must remain compliant. Be sure that only the standardized configurations are deployed and changes are properly approved and implemented by only the authorized personnel.

Ensure capacity optimization: Be prepared for current and future requirementswithout overbuying assets. To achieve this objective, youll have to estimate future capacity needs based on the current trends

Manage service performance: You must ensure that you meet commitments for performance levels as specified by your SLAs. It means striking a balance between the performance and resource utilization.