InteGrade: object‐oriented Grid middleware leveraging the idle computing power of desktop machines

Abstract

SUMMARYGrid computing technology improves the computing experiences at organizations by effectively integrating distributed computing resources. However, just a small fraction of currently available Grid infrastructures focuses on reutilization of existing commodity computing resources. This paper introduces InteGrade, a novel objectoriented middleware Grid infrastructure that focuses on leveraging the idle computing power of shared desktop machines. Its features include support for a wide range of parallel applications and mechanisms to assure that the owners of shared resources do not perceive any loss in the quality of service. A prototype implementation is under construction and the current version is available for download.

Generating Parallel Algorithms for Cluster and Grid Computing

“…The nodes are connected by a Switch 3COM 3300 FastEthernet 100Mb. We also used an 18-node grid using InteGrade (see [4]) middleware that provides efficient use of desktop microcomputers that are idle and available in a computer laboratory for graduate students, consisting of Pentium II 400 Mhz and 3 Athlon 1700+ desktop microcomputers interconnected in a local area network by 100Mb Ethernet. We have implemented and used some of the communication primitives of the Oxford BSPLib such as bsp put() and bsp get().…”

“…We illustrate the proposed method by deriving parallel algorithms for matrix multiplications and for computing the alignment score for string comparison. We then present implementation results on a 16-node Beowulf cluster and also on an 18-node grid, called InteGrade [4], consisting of desktop computers.…”

Scheduling Moldable BSP Tasks

“…Our group is developing a novel Grid middleware infrastructure called InteGrade [8]. The main principles of InteGrade are: modern object-oriented design, efficient communication based on CORBA, and native support for parallel computing.…”

“…In these new environments, the number of machines available for computation and their characteristics can change frequently. When we look at the case of Opportunistic Grid Computing, which uses the shared idle time of the existing computing infrastructure [8], the changes in machine availability occur even more rapidly. Thus, a model of parallel computation that does not allow variations in the number of processors available for computation would not fit well in this environment.…”

A Flexible Fault-Tolerance Mechanism for the Integrade Grid Middleware

“…The AutoGrid project uses the Integrade grid middleware [6] as the foundation for its implementation, incorporating autonomic mechanisms to its infrastructure in order to make its configuration and administration independent from human intervention. Our research focuses on adding to Integrade three autonomic properties: self-healing, self-optimization, and selfconfiguration.…”