Javier Echaiz

Extending an SSI Cluster for Resource Discovery in Grid Computing

Grid technologies enable large-scale sharing of resources within formal or informal consortia of individuals and/or virtual organizations. In these settings, the discovery, characterization, and monitoring of resources, services, and computations can be challenging due to the considerable diversity, large numbers, dynamic behavior, and geographical distribution of the entities in which a user might be interested. Hence, information services are a vital part of any grid software infrastructure, providing fundamental mechanisms for discovery and monitoring, and thus for planning and adapting application behavior. This paper proposes a resource discovery system for grid computing with fault-tolerant capabilities starting from an SSI clustering operating system. The proposed system uses dynamic leader-determination and registration mechanisms to automatically recover from nodes and network failures. The system is centralized and uses dynamic (or soft-state) registration to detect and recover from failures. Provisional or backup leader determination provides tolerance and recovery in the event of the leader node failing. The system was tested against a control network modeled after existing grid computing resource discovery components, such as Globus monitoring and discovery system (MDS). In various failure scenarios, the proposed system showed better resilience and performance than the control system

A Novel Algorithm for Indirect Reputation-Based Grid Resource Management

Component-based programming has been applied to address the requirements of applications in high performance computing (HPC). The usual service connectors of commercial component models do not fit some requirements of HPC, mainly regarding the support of parallelism, however. This paper looks at extensions to the usual notion of service connector to meet such requirements, using the # component model as a substratum, evidencing its expressiveness.A computational grid is a distributed infrastructure that appears to the end user as one large computing resource across organization boundaries. Grid technologies enable large-scale sharing of resources within formal or informal consortia of individuals and/or institutions, usually called virtual organizations. In these settings, the discovery, characterization, management, and monitoring of resources, services, and computations can be challenging due to the considerable diversity, large numbers, dynamic behavior, and geographical distribution of the entities in which a user might be interested. Trust is one of the biggest concerns in the grid resource management field. Grid systems can employ reputation mechanisms in order to provide this essential trust, but not usually without incurring in certain additional costs that negate the potential performance gains offered by grid computing technologies. Moreover, current reputation mechanisms are not appropriate for resource management in large-scale systems. In this paper, we present a new reputation model for resource management based on a economy model. Also we demonstrate how it can by employed to add trust into algorithms for grid scheduling. Finally, we simulate the proposed resource management algorithm in order to verify its effectiveness.

On Indirect Reputation-Based Grid Resource Management

Grid technologies enable large-scale sharing of resources within formal or informal consortia of individuals and/or institutions, usually called virtual organizations.In these settings, the discovery, characterization,management, and monitoring of resources,services, and computations can be challenging due to the considerable diversity, large numbers, dynamic behavior,and geographical distribution of the entities in which a user might be interested.Trust is one of the biggest concerns in the grid resource management field. Grid systems can employ reputation mechanisms in order to provide this essential trust, but not usually without incurring in certain additional costs that negate the potential performance gains offered by grid computing technologies. Moreover,current reputation mechanisms are not appropriate for resource management in large-scale systems (generally used in P2P).In this paper, we present a new reputation model for resource management based on a economy model, which can be employed to add trust into algorithms for grid scheduling. Finally, we simulate the proposed resource management algorithm in order to verify its effectiveness.

Latin American Perspectives on Grid Computing, from Bahia Blanca, Argentina

A computational grid is a distributed infrastructure that appears to the end user as one large computing resource across organization boundaries. Grid technologies enable large-scale sharing of resources within formal or informal consortia of individuals and/or institutions, usually called virtual organizations. In these settings, the discovery, characterization, and monitoring of resources, services, and computations can be challenging due to the considerable diversity, large numbers, dynamic behavior, and geographical distribution of the entities in which a user might be interested. This work explores the grid computing current state and perspectives in Latin America. In particular, we describe potential grid application areas and the experiences of our group in this promising field of research. Also, considering that grid trends are showing the importance of grid services we also identify some limitations and future research directions on this domain.