Detailed description of the thesis

The following is the detailed description of the goals of my thesis (as it can also be found on http://www.esp.win.ua.ac.be/projects/show/541):

IaaS clouds deliver infrastructure level-resources (processing power, memory, data storage, networking) as a commodity to clients through an on-demand and pay-per-use model [1]. At present, the model for trading cloud resources at the major providers (Amazon, Microsoft, Rightscale), involves the provider defining its own trading mechanism. Amazon for example [2], currently one of the most successful IaaS providers, sells its cloud resources through a spot market with prices that vary based on supply and demand during the week, through a reservation-based market with fixed prices, and through a fixed pricing scheme for non-reserved resources.

The commoditization of IT resources brought forth by the IaaS paradigm eases the deployment and management of applications on third party remote resources significantly. It also increases the possibilities for trading IaaS resources on a scale that moves beyond the individual provider level. Such inter-provider trading is used in other commodities markets such as electricity markets, to efficiently balance supply and demand on relatively short timescales. In addition, an open market for IaaS resources can efficiently open up the cloud IaaS provider space to less dedicated resources such as pools of desktop machines which represent a huge potential pool of processing power.

At present however, the possibilities for creating such open markets for IaaS resources have been largely unexplored. The goal of this thesis is to investigate: 1) Whether it is theoretically feasible to design an open market for cloud resources 2) The design options for open IaaS markets. In this respect, you will also look into the design of markets for energy trading and evaluate the differences and parallels with IaaS markets. 3) The behavior of the proposed market design in terms of price formation and stability through simulation

[1] M. Armbrust et. al, Above the Clouds: A Berkeley View of Cloud Computing, Armbrust et al., EECS Department, University of California, Berkeley, 2009, http://www.eecs.berkeley.edu/Pubs/TechRpts/2009/EECS-2009-28.html 

[2] Amazon, LLC, Amazon EC2 Pricing http://aws.amazon.com/ec2/pricing/ 

[3] N. Dube, Supercomputing futures the Next Sharing Paradigm for HPC Resources, http://www.nic.qc.ca/files/nicdubePhDthesis.pdf