Chapter Overview

Each of the technologies in HP's Virtual Server Environment provides environments with unique characteristics for hosting workloads. nPartitions, virtual partitions, virtual machines, and Secure Resource Partitions provide varying degrees of workload isolation, ranging from complete hardware isolation using nPartitions to resource isolation within an operating system using Secure Resource Partitions. In addition, each technology provides varying levels of granularity and flexibility with regard to resource allocation. The Utility Pricing strategies of HP's Virtual Server Environment, which include Instant Capacity and Temporary Instant Capacity, provide a means to purchase systems without initially activating all of the hardware, and pay per use provides an option to pay for only the resources that are actually consumed. Finally, using tools in the VSE Management Software allows workloads to be managed in a highly available and policy-based environment. As a result of the combined technologies in HP's Virtual Server Environment, many different choices and options are available for workload deployment in enterprise data centers.

While each of the VSE technologies fills a vital role in the overall solution, several questions can be difficult to answer without specialized tools. Questions such as, "If I consolidate these workloads, will the resulting system be able to handle the combined load?" and "How do I know how big to make my virtual machines?" are often asked but difficult to answer. Another question that can be perplexing is, "I'm deploying a new workload; where is the best place to put it?" In order to help capacity planners answer these types of questions and more, HP's Virtual Server Environment offers the Capacity Advisor product. Capacity Advisor automates the steps that are traditionally performed manually by capacity planners, thereby freeing the capacity planners to focus on planning for future workload introductions, migrations, and consolidations.

This chapter begins with an overview of Capacity Advisor. It describes a common Capacity Advisor use model and introduces the most commonly used terms relating to Capacity Advisor. It illustrates one of the many reasons for capacity planning by describing a planning scenario that was performed to determine whether two production workloads in HP's internal IT department could be consolidated. The chapter concludes with an example scenario that walks through a workload consolidation. In the consolidation example, three workloads running in separate nPartitions are evaluated to determine whether a single nPartition with fewer total resources is able to host the workloads. The goal of the scenario is to reduce the hardware necessary for the workloads while increasing resource utilization.

It should be noted this chapter was written using prototypes of the Capacity Advisor. The screens shown in this chapter will vary slightly from the final product.