Autonomic Malleability in Iterative MPI Applications

Ribeiro, Felipe Soares; Nascimento, Aline P.; Boeres, Cristina; Rebello, Vinod E. F.; Sena, Alexandre C.

doi:10.1109/sbac-pad.2013.4

Cited by 9 publications

(6 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In [63], the authors leveraged EasyGrid framework [64] to transform a traditional moldable MPI application into an autonomic malleable version that inherits the benefits of both malleability [12] and autonomy [65]. As such, applications will thus be capable of adjusting the number of their processes, if appropriate, and/or reallocating them to be adapted to the applications' requirements, resources availability, or resource failures in the system.…”

Section: Easygrid Application Management Systemmentioning

confidence: 99%

A Survey on Malleability Solutions for High-Performance Distributed Computing

et al. 2022

View full text Add to dashboard Cite

Maintaining a high rate of productivity, in terms of completed jobs per unit of time, in High-Performance Computing (HPC) facilities is a cornerstone in the next generation of exascale supercomputers. Process malleability is presented as a straightforward mechanism to address that issue. Nowadays, the vast majority of HPC facilities are intended for distributed-memory applications based on the Message Passing (MP) paradigm. For this reason, many efforts are based on the Message Passing Interface (MPI), the de facto standard programming model. Malleability aims to rescale executions on-the-fly, in other words, reconfigure the number and layout of processes in running applications. Process malleability involves resources reallocation within the HPC system, handling processes of the application, and redistributing data among those processes to resume the execution. This manuscript compiles how different frameworks address process malleability, their main features, their integration in resource management systems, and how they may be used in user codes. This paper is a detailed state-of-the-art devised as an entry point for researchers who are interested in process malleability.

show abstract

Section: Easygrid Application Management Systemmentioning

confidence: 99%

A Survey on Malleability Solutions for High-Performance Distributed Computing

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Baseados nesta mesma estratégia, duas abordagens para transformar aplicações iterativas MPI em aplicações maleáveis podem ser vistas em [El Maghraoui et al 2009] e em [Sena et al 2013]. Enquanto na primeira propostaé necessário chamar procedimentos para dividir/juntar processos para diminuir/aumentar a granularidade dos processos e, em seguida, migrar esses processos de acordo com os recursos disponíveis, na segunda proposta a aplicação MPIé transformada em uma aplicação autônoma, capaz de auto-ajustar a quantidade e granularidade de seus processos.…”

Section: Trabalhos Relacionadosunclassified

Ajuste Dinâmico de Threads para Execução Eficiente de Aplicações Iterativas OpenMP

Silva

Nascimento

Sena

2019

Anais Estendidos Do XX Simpósio Em Sistemas Computacionais De Alto Desempenho (SSCAD Estendido 2019)

View full text Add to dashboard Cite

Duas abordagens distintas podem ser adotadas para aumentar o desempenho das aplicações paralelas em uma arquitetura de alto desempenho: (i) ferramentas de auto-tuning; (ii) tornar ou criar aplicações que consigam se adaptar ao ambiente disponı́vel. Enquanto a primeira solução requer que o programador adapte sua aplicação a ferramenta e execute um profiling do ambiente, a segunda estratégia é mais complexa, pois necessita que o cientista seja capaz de criar aplicações com algum grau de autonomia. Assim, este trabalho propõe e avalia uma estrategia para ajustar dinamicamente a quantidade de theads de aplicações iterativas OpenMP. Os resultados obtidos mostram a viabilidade da estratégia proposta, que foi capaz de executar a aplicação para o problema de leilão eficientemente.

show abstract

“…Note that there are ways to reconfigure malleable MPI applications to use a different set of hardware resources. Ribeiro et al [39] propose a low-cost algorithm to transform traditional MPI applications into malleable versions that are capable of adjusting the number of running processes. This mechanism could be used to enhance our Controller.…”

Section: Controllermentioning

confidence: 99%

Cross-Layer Self-Adaptive/Self-Aware System Software for Exascale Systems

Gioiosa

Kestor

Kerbyson

et al. 2014

2014 IEEE 26th International Symposium on Computer Architecture and High Performance Computing

View full text Add to dashboard Cite

The extreme level of parallelism coupled with the limited available power budget expected in the exascale era brings unprecedented challenges that demand optimization of performance, power and resiliency in unison. Scalability on such systems is of paramount importance, while power and reliability issues may change the execution environment in which a parallel application runs. To solve these challenges exascale systems will require an introspective system software that combines system and application observations across all system stack layers with online feedback and adaptation mechanisms.In this paper we propose the design of a novel self-aware, selfadaptive system software in which a kernel-level Monitor, which continuously inspects the evolution of the target system through observation of Sensors, is combined with a user-level Controller, which reacts to changes in the execution environment, explores opportunities to increase performance, save power and adapts applications to new execution scenarios.We show that the monitoring system accurately monitors the evolution of parallel applications with a runtime overhead below 1-2%. As a test case, we design and implement a runtime system that aims at optimizing application's performance and system power consumption on complex hierarchical architectures. Our results show that our adaptive system reaches 98% of performance efficiency of manually-tuned applications.

show abstract

Autonomic Malleability in Iterative MPI Applications

Cited by 9 publications

References 23 publications

A Survey on Malleability Solutions for High-Performance Distributed Computing

A Survey on Malleability Solutions for High-Performance Distributed Computing

Ajuste Dinâmico de Threads para Execução Eficiente de Aplicações Iterativas OpenMP

Cross-Layer Self-Adaptive/Self-Aware System Software for Exascale Systems

Contact Info

Product

Resources

About