Resource Procurement in Cloud Systems
- Author:
- Kamrava, Sepideh
- Published:
- [University Park, Pennsylvania] : Pennsylvania State University, 2016.
- Physical Description:
- 1 electronic document
- Additional Creators:
- Urgaonkar, Bhuvan
- Access Online:
- etda.libraries.psu.edu
- Restrictions on Access:
- Open Access.
- Summary:
- With the growth in complexity of problems needed to be solved, the old fashionsystems cannot meet the users requirements. This niche has been identified bycloud providers offering enough resources with reasonable prices to handle anyrequests. The number of companies providing cloud services have been increasingin recent years. In order to compete with their rivals and attract more users, theyprovide more flexible options and plans. This dissertation focuses on one of thebiggest cloud resource providers, i.e., Amazon's EC2 and investigate its structureand introduce different resource types and plans in details. The cloud users tryto make an optimum decision on how they serve their demand with the availableoptions. This decision making can be really complex given the degrees of freedomthat exist in selecting among providers and their plans. Narrowing our attention ona smaller set of plans on Amazon's EC2 provider known as on-demand instances,reserved instances and the reserved instance marketplace, our goal is to derive a fewguidelines on how different methods of assigning demands to the available planscan affect the users'total cost. To decrease the service cost, some optimizationtechniques including integer linear programming and dynamic programming areutilized to derive an optimal or sub-optimal decision for many general workloads.The results illustrate that how adding more plans to the system can further reducethe cost when the demand is noisier while for pure periodic workloads, the optimalcost can be achieved by only using on-demand and reserved instances. Moreover,a threshold based decision method is proposed to determine when a user shouldsell their available reserved instances to achieve the maximum benefit. A heuristicmethod is further used to incorporate the migration cost as an instance of hiddencosts in the formulation. These techniques are applied to some real and syntheticworkloads to confirm the expectations and derive useful guidelines.
- Genre(s):
- Dissertation Note:
- M.S. Pennsylvania State University 2016.
- Technical Details:
- The full text of the dissertation is available as an Adobe Acrobat .pdf file ; Adobe Acrobat Reader required to view the file.
View MARC record | catkey: 19461298