Deadline-Driven Resource Management within Urgent Computing Cyberinfrastructure
نویسندگان
چکیده
One of the most important issues of urgent computing is the appropriate management of resources and tasks to meet the requirements of this concept. This article presents a model-based approach to a deadline-driven resource control taking into account the main requirements of urgent computing. The structure of stochastic multi-componential performance model for estimation and optimization of the task processing time is defined. The set of impact factors of the performance process including virtual resource management and services reliability is identified and modeled. The implementation of proposed approaches within urgent computing mode of CLAVIRE platform is described and illustrated with experiments. © 2013 The Authors. Published by Elsevier B.V. Selection and/or peer-review under responsibility of the organizers of the 2013 International Conference on Computational Science
منابع مشابه
The University of Chicago Deadline-based Grid Resource Selection for Urgent Computing a Thesis Submitted to the Faculty of the Division of the Physical Sciences in Candidacy for the Degree of Master of Science Department of Computer Science by Nick Trebon
Scientific simulation and modeling often aid in making critical decisions in such diverse fields as city planning, severe weather prediction, and influenza modeling. In some of these situations, the decisions must be made before a strict deadline, after which the results are of little use. Therefore, it is imperative that these urgent computations begin execution as quickly as possible. The Spe...
متن کاملDeadline and Budget Distribution based Cost- Time Optimization Workflow Scheduling Algorithm for Cloud
Cloud computing is a rapidly growing area. Cloud Computing offers utility-oriented IT services to the users worldwide over the internet. As compared to grid computing, the problem of resource management is transformed into resource virtualization and allocations. Effective scheduling is a key concern for the execution of performance driven applications, such as workflows in dynamic and cost-dri...
متن کاملLibra: An Economy driven Job Scheduling System for Clusters
Clusters of computers have emerged as mainstream parallel and distributed platforms for highperformance, high-throughput and high-availability computing. To enable effective resource management on clusters, numerous cluster managements systems and schedulers have been designed. However, their focus has essentially been on maximizing CPU performance, but not on improving the value of utility del...
متن کاملTowards Model Driven Verifiable Deployment of Distributed Simulations in Cloud
Running simulations on cloud computing platforms offers advantages to users such as on-demand computing and scalability. Despite these benefits offered by cloud computing platforms, limitations exists in verifiable and efficient deployment of these simulations in the cloud. Moreover, distributed simulations impose additional requirements of coordinated time stepped execution to progress the sim...
متن کاملTowards a Cyberaide Shell for the TeraGrid
Through the development of advanced middleware, Grid computing has evolved to a mature technology that scientists and researchers can leverage to gain knowledge that was previously unobtainable in a wide variety of scientific domains [2]. However, the ability for new scientists to become involved with Grid computing has become challenging due to its complex nature. This high entry barrier in tu...
متن کامل