A tool for the design and evaluation of hybrid scheduling algorithms for computational grids

Vianna, B. A.; Fonseca, A. A.; Moura, Nilmax Teones; Menezes, L. T.; Mendes, Hugo Abreu; Silva, J. A.; Boeres, Cristina; Rebello, Vinod E. F.

doi:10.1145/1028493.1028500

Cited by 3 publications

(5 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to space constraints the reader is referred to [19] for further details on the task and processor choice options available and the events which activate the dynamic scheduler.…”

Section: Dynamic Scheduling Strategiesmentioning

confidence: 99%

“…As discussed in Section 2.3.1, for static schedulers, the user can define one or more versions of the Configurable List Scheduling algorithm (see Figure 4) as well as use heuristics taken from the literature. In the case of hybrid schedulers, the user not only selects the static scheduler, but also the dynamic strategy and the scheduling event variant [19]. Given the specification of the scheduling strategy and its input data, the results can be evaluated graphically.…”

Section: Designing and Evaluating Scheduling Algorithmsmentioning

confidence: 99%

“…A table Algorithm versus DAG shows, for each graph, the makespan obtained and the number of processors required by the schedules produced by the respective heuristics. Also available is a table showing a pairwise comparison between the strategies under evaluation in terms of the number of worse, equal and better schedules, the percentage of times that each algorithm produces a schedule with the smallest makespan, and the overall quality of schedules produced relative to those of the other algorithms [19]. The designer can opt to save the resulting schedules, evaluations and Gantt charts for later reference.…”

Section: Designing and Evaluating Scheduling Algorithmsmentioning

confidence: 99%

See 2 more Smart Citations

An EasyGrid portal for scheduling system‐aware applications on computational Grids

Boeres

Fonseca

Mendes

et al. 2005

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

SUMMARYOne of the objectives of computational Grids is to offer applications the collective computational power of distributed but typically shared heterogeneous resources. Unfortunately, efficiently harnessing the performance potential of such systems (i.e. how and where applications should execute on the Grid) is a challenging endeavor due principally to the very distributed, shared and heterogeneous nature of the resources involved. A crucial step towards solving this problem is the need to identify both an appropriate scheduling model and scheduling algorithm(s). This paper presents a tool to aid the design and evaluation of scheduling policies suitable for efficient execution of system-aware parallel applications on computational Grids.

show abstract

“…Due to space constraints the reader is referred to [19] for further details on the task and processor choice options available and the events which activate the dynamic scheduler.…”

Section: Dynamic Scheduling Strategiesmentioning

confidence: 99%

Section: Designing and Evaluating Scheduling Algorithmsmentioning

confidence: 99%

Section: Designing and Evaluating Scheduling Algorithmsmentioning

confidence: 99%

See 1 more Smart Citation

An EasyGrid portal for scheduling system‐aware applications on computational Grids

Boeres

Fonseca

Mendes

et al. 2005

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

show abstract

“…This paradigm has been widely used to represent scientific processes [45]. With the emergence of Grid computing, the scheduling of workflows, particularly for e-Science applications, is receiving substantial attention [4,5,7,9,22,42,43,49]. Some known applications which make use of the workflow paradigm are Montage [18], AIRSN [54], LIGO [17], Chimera [3], and CSTEM [20], including applications in chemistry, biology, physics, and computer science.…”

Section: Introductionmentioning

confidence: 99%

Towards the Scheduling of Multiple Workflows on Computational Grids

Bittencourt

Madeira

2009

J Grid Computing

View full text Add to dashboard Cite

The workflow paradigm has become the standard to represent processes and their execution flows. With the evolution of e-Science, workflows are becoming larger and more computational demanding. Such e-Science necessities match with what computational Grids have to offer. Grids are shared distributed platforms which will eventually receive multiple requisitions to execute workflows. With this, there is a demand for a scheduler which deals with multiple workflows in the same set of resources, thus the development of multiple workflow scheduling algorithms is necessary. In this paper we describe four different initial strategies for scheduling multiple workflows on Grids and evaluate them in terms of schedule length and fairness. We present results for the initial schedule and for the makespan after the execution with external load. From the results we conclude that interleaving the workflows on the Grid leads to good average makespan and provides fairness when multiple workflows share the same set of resources.

show abstract

“…Esse paradigma tem sido usado amplamente na representação de processos científicos [93]. Com o surgimento das grades computacionais, o escalonamento de workflows, particularmente aqueles de aplicações de e-Ciência, têm recebido atenção substancial [5,40,6,80,12,77,97,27]. Algumas aplicações conhecidas que usam esse paradigma são Montage [37], AIRSN [108], LIGO [36], Chimera [3] e CSTEM [38], incluindo aplicações em química, biologia, física e ciência da computação.…”

Section: Grafos Acíclicos Direcionadosunclassified

Algoritmos para escalonamento de tarefas dependentes representadas por grafos acíclicos direcionados em grades computacionais

Bittencourt¹

View full text Add to dashboard Cite

Grades computacionais são sistemas distribuídos compartilhados potencialmente grandes compostos por recursos heterogêneos que são ligados através de uma rede com enlaces heterogêneos. Esses sistemas tornaram-se ambientes largamente difundidos para execução de tarefas que demandam grande capacidade de processamento. Por serem sistemas compartilhados, a submissão de tarefas nas grades é oriunda de diversos usuários independentemente, o que gera uma demanda concorrente pelos recursos computacionais que deve ser gerenciada pelo middleware da grade. O escalonador é o componente responsável por decidir de que forma a distribuição dessas tarefas será realizada, devendo tratar das peculiaridades desse ambiente, tais como a heterogeneidade e o comportamento dinâmico dos recursos que o compõem, com variações tanto em quantidade quanto em qualidade. A função objetivo mais comum encontrada no escalonamento de tarefas é a minimização do makespan, ou seja, o tempo de término das tarefas que estão sendo escalonadas. Dentre os possíveis tipos de tarefas executadas em grades podemos destacar as tarefas independentes, que executam sem comunicação entre si, e as tarefas dependentes, que possuem dependências de dados que geram precedências de execução e são freqüentemente modeladas como grafos acíclicos direcionados (DAGs -do inglês directed acyclic graphs). Dentre as aplicações compostas por tarefas dependentes, os DAGs de e-Ciência se sobressaem pela complexidade e necessidade crescente de recursos computacionais. Adicionalmente, o problema de escalonamento de tarefas, em sua forma geral, é NP-Completo. Dessa forma, o estudo do escalonamento de DAGs em grades computacionais é importante para o aprimoramento da execução de aplicações científicas utilizadas em diversas áreas do conhecimento.Nesta tese apresentamos algoritmos para quatro tipos de problema relacionados ao escalonamento de DAGs em grades: escalonamento estático de DAGs, escalonamento dinâmico de DAGs, escalonamento bi-critério e escalonamento de múltiplos DAGs. Apresentamos avaliações do makespan gerado pelos algoritmos após o escalonamento inicial e após a execução das tarefas com carga externa simulada nos recursos. vii

show abstract

A tool for the design and evaluation of hybrid scheduling algorithms for computational grids

Cited by 3 publications

References 14 publications

An EasyGrid portal for scheduling system‐aware applications on computational Grids

An EasyGrid portal for scheduling system‐aware applications on computational Grids

Towards the Scheduling of Multiple Workflows on Computational Grids

Algoritmos para escalonamento de tarefas dependentes representadas por grafos acíclicos direcionados em grades computacionais

Contact Info

Product

Resources

About