Engineering an autonomic partitioning framework for Grid-based SAMR applications
High performance scientific and engineering computing
Autonomic Web-Based Simulation
ANSS '05 Proceedings of the 38th annual Symposium on Simulation
A software architecture approach for structuring autonomic systems
DEAS '05 Proceedings of the 2005 workshop on Design and evolution of autonomic application software
AutoMate: Enabling Autonomic Applications on the Grid
Cluster Computing
SpringSim '07 Proceedings of the 2007 spring simulation multiconference - Volume 2
Operations Research Methods for Autonomic Resource Management
Journal of Network and Systems Management
Enabling autonomic grid applications: requirements, models and infrastructure
Self-star Properties in Complex Information Systems
| Hi-index | 0.00 |
With rapid technological advances in networkinfrastructure, programming languages, compatiblecomponent interfaces and so many more areas, today thecomputational Grid has evolved with the potential ofseamless aggregation, integration and interactions. Thishas made it possible to conceive a new generation ofrealistic, scientific and engineering simulations ofcomplex physical phenomenon. These applications willsymbiotically and opportunistically combinecomputations, experiments, observations, and real-timedata, and will provide important insights into complexphenomena. However, the phenomena being modeled areinherently complex, multi-phased, multi-scaled, dynamicand heterogeneous (in time space and state).Furthermore, their implementations involve multipleresearchers with scores of models, hundreds ofcomponents and dynamic compositions and interactionsbetween these components. The underlying Gridinfrastructure is similarly heterogeneous and dynamic,globally aggregating large numbers of independentcomputing and communication resources, data stores andsensor networks. The combination of the two results inapplication development, configuration and managementcomplexities that break current paradigms that are basedon passive components and static compositions. In fact,we have reached a level of complexity, heterogeneity, anddynamism that our programming environments andinfrastructure are becoming unmanageable/insecure [3].In this paper we attempt to explore an alternativeprogramming paradigm and management technique thatis based on strategies used by biological systems to dealwith complexity, heterogeneity and uncertainty. Thisapproach is referred to as autonomic computing [1]. Wediscuss key technologies to enable the development ofautonomic Grid applications. We also present amiddleware architecture that sits on top of the existingGrid middleware, intelligently managing and executingautonomic applications with huge computationalrequirements over limited Grid resources. We discuss indetail how the proposed vGrid middleware can be used todynamically control and manage large-scale forest firesimulation.