Job Dispatching and Scheduling under Heterogeneous Clusters
Date Issued
2014
Date
2014
Author(s)
Lin, Ting-Chou
Abstract
Many enterprises or institutes are building private clouds by establishing their own data center. In such data centers, the physical machines can be different due to annual upgrades, but the amount of machine are fixed for most of the time. In such heterogeneous environment, scheduling jobs with different resource requirements and characteristic in order to meet different timing constraints is important. In this paper, we proposed a cloud resource management framework to dynamically adjust the number of computation nodes for every job in the system.
Subjects
Job/Task Scheduling
Heterogeneous Clusters
Resource Allocation
Type
thesis