DOI: 10.5176/978-981-08-5837-7_183
Authors: Markus Winter
Abstract:
Over the past years, the operation of data centers became a major challenge for application service providers and corporate IT departments. A growing number of heterogeneous system landscapes leads to rising costs for infrastructure operations and management. The increasing demand for consolidation of computing resources within the data center is a reaction to the continuous growth of underutilized server units that has led to an inefficient utilization of hardware infrastructure as a whole. This paper presents the findings of a case study on resource utilization of very large business applications (VLBA). It is based on collected utilization data from about 700 physical servers from a large data center provider containing multiple different customer environments that were selected for detailed analysis. The case study validates oversizing statements based on real-world data center workloads and identifies potential consolidation parameters for improved infrastructure sizing. It analyzes the utilization of CPU, RAM, network I/O, and disk I/O with regard to a potential consolidation of server infrastructure. The results of the study show that CPU utilization is not necessarily the only indication for oversizing and hence a holistic approach is required to enable reliable initial sizing estimations for a consolidated data center infrastructure. In addition, the study presents an outlook beyond server consolidation endeavors and discusses the transformation path of corporate data centers into compute clouds. Consolidation based on a thorough analysis of existing infrastructure utilization and sizing is an essential step before shifting any software environments into infrastructure clouds, especially with regard to VLBAs. This is key for tangible benefits and efficient IT utilization in infrastructure as a service (IaaS) cloud models.
Updating...