Kaizar Amin

GridAnt: a client-controllable grid workflow system

Process management is an extremely important concept in both business and scientific communities. Several workflow management tools have been proposed in recent years offering advanced functionality in various domains. In the business world, workflow vendors offer commercial and customized solutions targeting specific users. In the scientific world, several open-source workflow management tools are freely available. However they are directed toward service aggregation rather than distributed process management. Little consideration is given to the needs of the client in terms of mapping the process flow of the client. In the grid community it is essential that the grid users have such a tool available enabling them to orchestrate complex work-flows on the fly without substantial help from the service providers. At the same time it is important that the grid user not be burdened with the intricacies of the workflow system. With the perspective of the grid user in mind, an extensible client-side workflow management system, called GridAnt, has been developed. This paper discusses the design principles, functionality, and application of the proposed GridAnt workflow manager.

Analysis and Provision of QoS for Distributed Grid Applications

Abstract Grid computing provides the infrastructure necessary to access and use distributed resources as part of virtual organizations. When used in this way, Grid computing makes it possible for users to participate in collaborative and distributed applications such as tele-immersion, visualization, and computational simulation. Some of these applications operate in a collaborative mode, requiring data to be stored and delivered in a timely manner. This class of applications must adhere to stringent real-time constraints and Quality-of-Service (QoS) requirements. A QoS management approach is therefore required to orchestrate and guarantee the timely interaction between such applications and services. We discuss the design and a prototype implementation of a QoS system, and demonstrate how we enable Grid applications to become QoS compliant. We validate this approach through a case study of an image processing task derived from a nanoscale structures application.

An OGSA-Based Quality of Service Framework

Grid computing provides a robust paradigm to aggregate disparate resources in a secure and controlled environment. Grid architectures require an underpinning Quality of Service (QoS) support in order to manage complex data and computation intensive applications. However, QoS guarantees in the Grid context have not been given the attention they merit. In order to enhance the functionality offered by computational Grids, we overlay the Grid framework with an advanced QoS architecture, called G-QoSM. The G-QoSM framework provides a new service-oriented QoS management model that leverages the Open Grid Service Architecture (OGSA) and has a number of interesting features: (1) Grid service discovery based on QoS attributes, (2) policy-based admission control for advance reservation support, and (3) Grid service execution with QoS constraints. This paper discusses the different components of the G-QoSM framework, in the context of OGSA architectures.

QoS support for high-performance scientific Grid applications

The Grid approach provides the ability to access and use distributed resources as part of virtual organizations. The emerging Grid infrastructure gives rise to a class of scientific applications and services to support collaborative and distributed resource-sharing requirements as part of teleimmersion, visualization, and simulation services. Because such applications operate in a collaborative mode, data must be stored and delivered in timely manner to meet deadlines. Hence, this class of applications has stringent real-time constraints and quality-of-service (QoS) requirements. A QoS management approach is required to orchestrate and guarantee the interaction between such applications and services. In this paper we discuss the design and prototype implementation of a QoS system and show how we enable Grid applications to become QoS compliant. We validate this approach through a case study of nanomaterials. Our approach enhances the current Open Grid Services Architecture. We demonstrate the usefulness of the approach on a nanomaterials application.

Agent-Based Distance Vector Routing

Mobile Agents are being proposed for an increasing variety of applications. Distance Vector Routing (DVR) is an example of one application that can benefit from an agent-based approach. DVR algorithms, such as RIP, have been shown to cause considerable network resource overhead due to the large number of messages generated at each host/router throughout the route update process. Many of these messages are wasteful since they do not contribute to the route discovery process. However, in an agent-based solution, the number of messages is bounded by the number of agents in the system. In this paper, we present an agent-based solution to DVR. In addition, we will describe agent migration strategies that improve the performance of the route discovery process, namely Random Walk and Structured Walk.

Agent-based distance vector routing: a resource efficient and scalable approach to routing in large communication networks

In spite of the ever-increasing availability of computation and communication resources in modern networks, the overhead associated with network management protocols, such as traffic control and routing, continues to be an important aspect in the design of new methodologies. Resource efficiency of such protocols has become even more prominent with the recent developments of wireless and ad hoc networks, which are marked by much more severe resource constraints in terms of bandwidth, memory, and computational capabilities. This paper presents an agent-based approach to distance vector routing (ADVR) that addresses these resource constraints. ADVR is a resource efficient implementation of distance vector routing that is fault tolerant and scales well for large networks. ADVR draws upon some basic biologically inspired principles to facilitate coordination among the mobile agents that implement the routing task. Specifically, simulated pheromones are used to control the movement of agents within the network and to dynamically adjust the number of agents in the population. The behavior of ADVR is analyzed and compared to that of traditional distance vector routing.

Grid Computing for the Masses: An Overview

The common goals of the Grid and peer-to-peer communities have brought them in close proximity. Both the technologies overlay a collaborative resource-sharing infrastructure on existing (public) networks. In realizing this shared goal, however, they concentrate on significantly contrasting issues. The Grid paradigm focuses on performance, control, security, specialization, and standardization. On the other hand, the peer-to-peer paradigm concentrates on fault tolerance, resilience, decentralization, and peer cooperation. In this paper, we discuss Grid usage models including traditional Grids, ad hoc Grids, and federated Grids. We compare these approaches to peer-to-peer computing and discuss the issues involved in the convergence of the two paradigms.

A Grid Service-Based Active Thermochemical Table Framework

In this paper we report ourwork on the integration of existing scientific applications using Grid Services. We describe a general architecture that provides access to these applications via Web services-based application factories. Furthermore, we demonstrate how such services can interact with each other. These interactions enable a level of integration that assists the scientific application architect in leveraging applications running in heterogeneous runtime environments. Our architecture is implemented by using existing infrastructures and middleware, such as Web services, the Globus Toolkit, and the Java CoG Kit. We test our architecture on a thermochemistry application that provides a number of requirements, such as batch processing, interactive and collaborative steering, use of multiple platforms, visualization through large displays, and access via a portal framework. Besides the innovative use of the Grid and Web services, we have also provided a novel algorithmic contribution to scientific disciplines that use thermochemical tables. Specifically, we modified the original approach to constructing thermochemical tables to include an iterative process of refinement leading to increased accuracy; we are now implementing this approach. We have designed a portal for accessing the set of services provided, which include the display of network dependencies between the reactions a chemist may be interested in and interactive querying of associated species data.

A Framework for Building Scientific Knowledge Grids Applied to Thermochemical Tables

An important part of enabling the future information infrastructure is to provide convenient services for application users that make such an environment highly functional and easy to use for the non-expert. In other words, the scientist should be able to focus on the science but should have sophisticated tools at hand that enable new and enhanced modalities in which science can be performed. We describe a general architecture that provides access to scientific applications. We demonstrate how such services can interact with each other and be reused by the scientific application architect to orchestrate Grid services. Our architecture is implemented by reusing concepts, infrastructures, and middleware from Web and Grid services, as well as from the newest Globus Toolkit and the Java CoG Kit. We apply our general architecture to a specific application: Active Thermochemical Tables. This application demonstrates a number of interesting and future-oriented uses, such as the need for batch processing, interactive and collaborative steering, use of multiple platforms,visualization through large displays, and access via a portal framework. Besides the innovative use of the Grid and Web services, we also provide a novel algorithmic contribution to scientific disciplines that use thermochemical tables. Specifically, we modified the original approach to constructing thermochemical tables to include an iterative process of refinement leading to increased accuracy. We have designed a variety of access environments including a shell, a Grid desktop, and a portal for accessing the set of services provided, which include the display of network dependencies between the reactions a chemist may be interested in and interactive querying of associated species data. Modalities that need to be supported are the protection of intellectual property, the creation of metadata and information shared with and gathered by the community, and easy access to the service by the application scientists.

Ad hoc grid security infrastructure

This paper describes an ad hoc grid security infrastructure developed as a part of the Java CoG Kit project. It supports several requirements specific to the sporadic nature of ad hoc grids. It focuses on identity management, identity verification, and authorization control in spontaneous grid collaborations without pre-established policies or environments. It adopts established community standards, with modifications where needed. This paper also discusses the integration of the ad hoc grid security infrastructure in an ad hoc grid implementation. The implementation supports secure collaboration in ad hoc grids using commodity technologies such as the Java CoG Kit, JXTA, GSI, and XACML.