GriT: a CORBA-based grid middleware architecture

Abstract

Rapid advances in networking, hardware, and middleware technologies are facilitating the development and deployment of grid applications, which are characterized, by their very high computing and resource requirements. These applications and services have multiple, simultaneous end-to-end quality of service (QoS) requirements, such as delay guarantees, jitter guarantees, security, scalability, reliability and availability guarantees, and bandwidth and throughput guarantees. Moreover, these applications and services require secure, controlled, reliable, and guaranteed access to different types of resources, such as network bandwidth, computing power, and storage capabilities, available from multiple service providers. To support next-generation grid applications effectively, there is a need to simplify grid programming by developing a new grid middleware that raises the level of abstraction, and reduces the accidental complexities incurred, by programming at the grid infrastructure middleware level offered by existing grid middleware such as Globus, ICENI, and Legion. Moreover, the new grid middleware must ensure multiple end-to-end QoS properties simultaneously. The paper provides three contributions to the research on next generation grid middleware architecture that provides the above-mentioned properties. First, we describe how we are utilizing the standards-based CORBA distributed object computing and integration technology to design the next generation grid middleware, called Grid TAO (GriT) that complements and enhances existing low-level Grid middleware, such as Globus. Second, we describe how we are using the real-time, fault-tolerant, and data parallel CORBA features to implement GriT to provide the desired properties. Finally, we show how CORBA's platform and language independence features are used in GriT to resolve the portability and interoperability challenges faced, by current grid applications.