Grid Computing Infrastructure Architecture Document

GSELK VISION

Note ==> This document does not discuss an implementation view, the purpose of the “Grid Enabled Secure Linux Kernel's” Consortium is to develop standardized domain schema, message schema and service definitions not to deliver working implementations.
However; it is important that we clearly delineate the definition of the domain concerns from the service concerns.

This will become clearer in the elaboration below, but at the heart we need to separate the common, domain knowledge from the technical infrastructure and constraints on implementation. Although we have decided that the realization of this architecture will be based on various self defined services there is always the possibility that in the future such technical infrastructures change. We will also see that by separating these concerns in this way we can handle technical details such as transportation, Instructiveness, Scalability, Fault Recovery, Discovery, Access, Usability, Hierarchy, Status, Dynamics, Automatic Adaptation, security and privacy concerns in a clean fashion.

3 Architectural Goals and Constraints

The key concerned areas of Distributed and parallel computing and an effort to standardize the Grid Protocols have driven the explosive research into the filed of Grid, Mata and High performance computing. For the current and future phases of growth in technology and its limitations, the industry and organizations will use the Grid technology to fulfill their thrust for high end computing and massive computational demands. These current research efforts in this regard will streamline the development process for creating a highly interactive and inter-operable suite of software for high end computational needs and our work will hopefully contribute for the communities striving for advancement and perfection in the fields of Kernel optimization and Grid computing i.e. Not only facilitate and enhance the merging of Grid technology and operating System but also provide the built-in support at kernel level for this technological challenge which in turn will help in the computational challenges.

With this in mind we may envision an end-result as seen from a human user similar to the diagram below, alternate views are possible depending on the role of the client; it may be the 1^st, 2^nd or nth client requiring computational services.