Abstract
Distributed virtual machines can help to build scalable, manageable, and efficient grid infrastructures. The work proposed in this paper focuses on employing virtual machines for grid computing. In order to efficiently run grid applications, virtual machine resource information should be provided. This paper first discusses the system architecture of virtual machine pools and the process of information retrieval from virtual machines. Based on the characterization of the system model, this paper presents the work on how to retrieve resource information from Xen/VMware virtual machines via VMware Common Information Model Software Development Kit and lightweight Java agents. The resource information is integrated into a grid information service. The work is implemented in a test bed with Xen/VMware virtual machines and the Globus Toolkit. With a performance evaluation and discussion on a real test bed, it is declared that the design and implementation of information services for virtual-machine-based grid systems are feasible, efficient, and scalable.
Original language | English |
---|---|
Pages (from-to) | 1362-1374 |
Number of pages | 13 |
Journal | IEEE Transactions on Systems, Man and Cybernetics: Systems |
Volume | 40 |
Issue number | 6 |
DOIs | |
Publication status | Published - 1 Nov 2010 |
Keywords
- Grid computing
- virtual machine
- information service