Font Size: a A A

The Design And Implementation Of A Service Oriented Grid Architecture

Posted on:2009-11-08Degree:MasterType:Thesis
Country:ChinaCandidate:S P XuFull Text:PDF
GTID:2178360272991978Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Although the power of computers to process data has progressed at an astounding rate over the past decade, the problems to which people wish to apply this compute power have exploded in magnitude as well. A common approach to solving a demanding compute problem is to employ an algorithm that allows multiple processors to work on the problem in parallel, and to run this algorithm on either a symmetric multiprocessing (SMP) computer, or more recently, on a compute grid that harnesses the collective processing power of multiple computers distributed over a network. The readiness, low-cost availability of high network bandwidth and networkable CPUs, coupled with improvements in software networking technologies, has enabled lower cost compute grids today. The ability of such compute grids to easily and affordably scale to meet the needs of large scale computation problems makes this approach to parallel processing appealing to a wide range of users today.Starting with the concept and origin of grid technology, this paper does in-depth analysis and applied research about, mainstream grid projects, scheduling algorithm in grid, service-oriented distributed architecture, and Jini technology. Based on this, service-oriented grid architecture which consists of a series of services with capability of plug-and-play, self-discovery, self-organizing and self-curing is designed in the paper. Grid system with this architecture has high reliability and can run with less monitoring and administrating. The architecture removes some limitations of the one built based on Web service. A more flexible method of resource definition than the one built based on Web service is harnessed. Moreover, idea of Boolean resource definition which is a simple and efficient solution to simplify mapping between job and resource is provided. The Task execution service in the architecture can dynamically report change of attributes to sent-out proxy and service register center through remote event, which can dramatically maximize accuracy of job scheduling.The open source distributed workflow engine, Pegasus is extended in the system. And resource description is enhanced to be more fine-granted. The task scheduler itself is abstracted as a service in this architecture. Idea of loading balance and key data synchronization between multiple task schedulers is employed in the architecture.Reliability design for the service-oriented grid system is discussed in the paper. And the leasing mechanism is harnessed in the architecture.How to implement SSO (Single Sign On) in the heterogeneous grid environment is well discussed in the paper. And a effective solution to solve the problem that the grid system employs SSL for data transmission, but most users do not have private key and certificate to sign on is provided.Finally, taking the implementation of the task execution service as an example, this paper introduces how to implement a Jini-based service with capability of self-discovery, self-organizing, and self-curing. With a fine-designed task, functionalities and performance has been tested for a long time. And testing reports have been provided.
Keywords/Search Tags:Grid Computing, SOA, Resource Management, Task scheduling
PDF Full Text Request
Related items