This paper examines the critical importance of capacity planning and performance monitoring as enterprises migrate computing resources from on-premise infrastructure to cloud-based platforms. The analysis focuses on virtualization as the foundational technology enabling scalable, efficient cloud deployments, with reference to major providers including Amazon Web Services, VMware, IBM, Microsoft, and Oracle. The paper discusses Infrastructure-as-a-Service (IaaS) and Platform-as-a-Service (PaaS) models, single- versus multitenancy strategies, and the role of business intelligence in monitoring cloud performance. Amazon's evolution from e-commerce platform to cloud services provider is used to illustrate how virtualization can drive entirely new business models.
The paper effectively uses a case-study-style example (Amazon's transition from e-commerce operator to AWS provider) as an extended illustration of a theoretical concept — virtualization as a "force multiplier." This technique of anchoring technical argument in a recognizable organizational narrative helps readers connect infrastructure concepts to strategic business outcomes.
The paper opens with a problem statement establishing why cloud migration demands careful capacity planning. The body section surveys virtualization technologies and their role in IaaS/PaaS models, weaving in vendor examples (VMware, IBM, Oracle, Amazon). A brief discussion of business intelligence ties monitoring practices back to the opening argument. The conclusion restates the stakes concisely. The structure is linear and deductive, moving from general need to specific technologies to practical implications.
The need for computer capacity evaluation and continual monitoring becomes even more critical when an enterprise is transferring its computing power to cloud-based systems. This transition from on-premise to cloud-based computing introduces exponentially greater levels of variables and performance issues that can have an immediate and significant effect on an enterprise's ability to perform even the most fundamental IT tasks (Read, 2010). Computing power transitioned to the cloud must be managed with a deliberate strategy that ensures a high degree of scalability, security, and a clear approach to single- versus multitenancy — that is, the allocation of specific dedicated memory and processor resources for specific tasks (Rasin, 2010). The intent of this analysis is to evaluate those areas of cloud-based systems' computing resources and define how enterprises can safeguard their applications and performance over the long term.
The catalyst of any successful transition of computing power from on-premise enterprise systems to cloud-based solutions is the concept of virtualization. IBM, Microsoft, Oracle, and VMware are all multibillion-dollar companies creating sophisticated technologies for virtualized server architectures that ensure the highest levels of computing performance possible for a given enterprise usage scenario. Virtualization is a series of technologies that enable a traditional server to be optimized for the specific needs of an enterprise-level software deployment (Luo, 2010). One example of virtualization being successfully used to scale computing power in the cloud is Microsoft's CRM system, which runs in partitioned, multitenant configurations within their own data centers.
Virtualization is also the core technology underlying Infrastructure-as-a-Service (IaaS) and Platform-as-a-Service (PaaS) cloud computing components, and it is an area of intensive investment by Amazon through their Web Services (AWS) platform (Engebretson, 2011). Virtualization's role in enabling cloud computing is a core aspect of the economics of this nascent yet rapidly growing area of technology. Implicit in the transfer of applications from on-premise to the cloud is the need for replicable server performance — in effect, creating a virtual server (Luo, 2010).
This is precisely what the virtualization technologies that Amazon is developing from a Research & Development (R&D) standpoint are designed to achieve: parsing literally thousands of servers into virtual machines that can be leased to enterprise customers (Engebretson, 2011). Virtualization requires the precise differentiation of processor power and cycles, allocated in an optimized sequence to ensure the highest performance possible. This is why cloud computing is still an imprecise science at best, with significant performance lags and inconsistencies on large-scale workloads (Luo, 2010). VMware is concentrating on these performance aberrations to create an optimized product and services mix across its software and hardware portfolio, aiming to ensure high levels of computer capacity utilization over time.
Amazon Web Services drew directly on lessons learned from managing one of the most successful e-commerce platforms in the world. Those lessons proved directly applicable to launching AWS, which relies heavily on virtualization at its foundational level (Engebretson, 2011). Scalability, for Amazon, ultimately meant supporting over fifty websites and thousands of transactions per second worldwide in multiple currencies. The lessons learned in building what became the first cloud-based distributed order management system gave Amazon insights into how virtualization could transform millions of other businesses as well (Engebretson, 2011). Embedded in the story of Amazon's move into offering AWS as a business are insights into how virtualization can serve as a force multiplier in the development of entirely new business models.
Cloud computing is changing the economics of enterprise computing (Luo, 2010). The continual transfer of computing capacity from on-premise to hosted platforms will continue, further underscoring the need for greater monitoring of cloud-based processor and system performance (Read, 2010). The bottom line is that cloud computing performance will make or break entire companies today and in the future — hence the need to monitor it closely and continuously.
Engebretson, J. (2011). Level 3 offers direct connectivity to Amazon Web Services. Connected Planet, n/a.
Luo, Y. (2010). Network I/O virtualization for cloud computing. IT Professional Magazine, 12(5), 36–41.
Rasin, A. (2010). MapReduce and parallel DBMSs: Friends or foes? Communications of the ACM, 53(1), 64.
Read, B. (2010). Buying decisions 2011: On-premises or hosted. Customer Inter@ction Solutions, 29(7), 6–9.
Shen, G. (2011). Cloud computing — the catalyst for self-service BI. Information Management, 21(5), 20.
Always verify citation format against your institution’s current style guide requirements.