Maturing virtualization in information technology systems has enabled increased implementations of the cloud com-puting paradigm, dissolving the need to co-locate user and computing power by providing desired services through the network. This thesis researches the support that current network modeling and simulation applications can provide to IT projects in planning, implementing and maintaining networks for cloud solutions. A problem-appropriate do-main model and subsequent requirements are developed for the assessment of several network modeling and simula-tion tools, which leads to the identification of a capability gap precluding the use of such tools in early stages of cloud computing projects. Consequently, a practical, modular designed methodology is proposed to measure the essential properties necessary for developing appropriate cloud computing network traffic models. The conducted proof-of-concept experiment applied to a virtual desktop environment finds the proposed methodology suitable and problem-appropriate, and results in recommended steps to close the identified capability gap.
http://archive.org/details/infrastructuresu109455580
Outstanding Thesis
Approved for public release; distribution is unlimited.