A Case For Grid Computing On Virtual Machines
ICDCS '03 Proceedings of the 23rd International Conference on Distributed Computing Systems
The Anatomy of the Grid: Enabling Scalable Virtual Organizations
International Journal of High Performance Computing Applications
Deploying virtual machines as sandboxes for the grid
WORLDS'05 Proceedings of the 2nd conference on Real, Large Distributed Systems - Volume 2
Towards virtual networks for virtual machine grid computing
VM'04 Proceedings of the 3rd conference on Virtual Machine Research And Technology Symposium - Volume 3
OCALA: an architecture for supporting legacy applications over overlays
NSDI'06 Proceedings of the 3rd conference on Networked Systems Design & Implementation - Volume 3
VTDC '06 Proceedings of the 2nd International Workshop on Virtualization Technology in Distributed Computing
Characterizing User-Level Network Virtualization: Performance, Overheads and Limits
ESCIENCE '08 Proceedings of the 2008 Fourth IEEE International Conference on eScience
IP over P2P: enabling self-configuring virtual IP networks for grid computing
IPDPS'06 Proceedings of the 20th international conference on Parallel and distributed processing
On the design of scalable, self-configuring virtual networks
Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis
Dynamic scheduling of virtual machines running HPC workloads in scientific grids
NTMS'09 Proceedings of the 3rd international conference on New technologies, mobility and security
Hi-index | 0.02 |
The rising popularity of grid computing makes the issues of growth, security, and access critical in deploying and maintaining well-functioning grid systems. Overlay networks (ON)provide a framework to deal with these issues, but current techniques impose limitations and administrative burdens such as manual configuration for each new system in the grid, installation and configuration of software. Additionally, current approaches lack methods of effectively merging clusters with individual workstations, usually focusing on either the grouping of distributed clusters or a desktop/workstation Grid. The main difference between the two scenarios is that in a cluster environment all machines share a common ON router, whereas in a workstation environment each machine has ON software. This paper presents a novel approach of self-configuring IP-based Virtual Private Networks (VPNs)1 overlays that support dynamic, seamless addition of new resources to the grid for both cluster and workstation platforms. The approach allows for bridging physical and virtual networking in clusters, in a manner that allows dynamic configuration of IP addresses while avoiding overlay routing among nodes within the same layer 2 network. To enable these features, the ON runs on top of a Peer-To-Peer (P2P) network that provides supports a distributed data store. IP addresses are dynamically allocated by a virtual DHCP server controlled by the ON router through atomic operations on the distributed data store. This atomic operation creates a mapping of an IP address to a P2P address that can later be used by the VPN and router to determine the host of an IP address. We have prototyped this approach, demonstrating the ability to seamlessly mix both workstation and cluster based approaches into a wide-area Condor pool.