Managing the Execution of Large Scale MPI Applications on Computational Grids

Nascimento, Ad.P.; Sena, Alexandre C.; Silva, J.A. da; Vianna, D.; Boeres, Cristina; Rebello, Vinod E. F.

doi:10.1109/cahpc.2005.30

Cited by 12 publications

(19 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In some instances, performance benefits exist for executing application processes concurrently. The optimal number depends on the size and I/O characteristics of the processes [5] and the workload of the machine. On the other hand, if the application is to run as a background job, a policy might be employed to only allow processes to execute when the machine is idle.…”

Section: The Hdsmentioning

confidence: 99%

“…Each layer or subsystem performs a specific service essential to the MPI application's autonomic behaviour and efficient execution in Grid environments. Based on the monitoring information provided, this architecture permits a higher level subsystem to modify the behaviour of a lower level subsystem [5].…”

Section: Easygrid Ams Middlewarementioning

confidence: 99%

“…How this information is collected and used by the dynamic scheduling subsystem is described in detail in the next section. An overview of the fault-tolerance layer, responsible for the re-execution of processes that fail, can be found in [5].…”

Section: Easygrid Ams Middlewarementioning

confidence: 99%

“…The AMS is embedded automatically into a user's parallel MPI application without modification of the original code by a scheduling portal [3] which, as with almost all management systems, also leverages the core middleware functionalities offered by the Globus Toolkit [4] to facilitate the use of Grid computing. The implementation of an AMS and the viability of transforming legacy clusterbased MPI applications into system-aware applications capable of executing in Grid environments was discussed in [5]. The AMS's subsumption architecture and hierarchical management structure designed to provide non-intrusive monitoring, scheduling and fault-tolerance support for Smart G-Apps (systemaware applications) were also introduced.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Distributed and dynamic self‐scheduling of parallel MPI Grid applications

Nascimento

Sena

Boeres

et al. 2006

Concurrency and Computation

View full text Add to dashboard Cite

SUMMARYThe execution of distributed applications on the Grid is already a reality. However, as both the number of applications grow and Grids increase in scale, the efficient utilization of the available but shared heterogeneous resources will become increasingly essential to the Grid's successful maturity. Furthermore, it is unclear whether existing Grid management systems are capable of meeting this challenge. The EasyGrid middleware is a hierarchically distributed application management system (AMS) that is embedded into MPI applications to autonomously orchestrate their execution efficiently in computational Grids. The overhead of employing a distinct AMS to make each application system aware brings at least two benefits. First, the adopted policies can be tailored to the specific needs of each application, leading to improved performance. Second, distributing the management effort among executing applications makes Grid management more scalable. This article focuses on scheduling policies of an AMS for a particular class of application, describing a low intrusion implementation of a hybrid scheduling strategy designed to elicit good performance even in dynamic environments such as Grids. Using application-specific scheduling policies, near-optimal runtimes highlight the advantages of self-scheduling when executing one or more system aware applications on a Grid.

show abstract

Section: The Hdsmentioning

confidence: 99%

Section: Easygrid Ams Middlewarementioning

confidence: 99%

Section: Easygrid Ams Middlewarementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Distributed and dynamic self‐scheduling of parallel MPI Grid applications

Nascimento

Sena

Boeres

et al. 2006

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…The EasyGrid Application Management System (AMS) transforms clusterbased MPI applications (designed for homogeneous, stable environments) into autonomic ones capable executing robustly and efficiently in Grids [3]. The EasyGrid AMS is an application-specific application-level middleware that is embedded into the user's MPI program at compile time, thus creating a self-contained autonomic MPI application.…”

Section: Introductionmentioning

confidence: 99%

Low Cost Self-healing in MPI Applications

Silva

Rebello

2007

Recent Advances in Parallel Virtual Machine and Message Passing Interface

View full text Add to dashboard Cite

Abstract. Writing applications capable of executing efficiently in Grids is extremely difficult and tedious for inexperienced users. The distributed resources are typically heterogeneous, non-dedicated, and are offered without any performance or availability guarantees. Systems capable of adapting the execution of an application to the dynamic characteristics of the Grid are essential. This work describes the strategy used to bestow the self-healing property on autonomic EasyGrid MPI applications to withstand process and resource failures. This paper highlights both the difficulties and the low cost solution adopted to offer fault tolerance in applications based on the standard Grid installation of LAM/MPI.

show abstract

A taxonomy and survey on autonomic management of applications in grid computing environments

Rahman

Ranjan

Buyya

et al. 2011

Concurrency and Computation

View full text Add to dashboard Cite

SUMMARYIn Grid computing environments, the availability, performance, and state of resources, applications, services, and data undergo continuous changes during the life cycle of an application. Uncertainty is a fact in Grid environments, which is triggered by multiple factors, including: (1) failures, (2) dynamism, (3) incomplete global knowledge, and (4) heterogeneity. Unfortunately, the existing Grid management methods, tools, and application composition techniques are inadequate to handle these resource, application and environment behaviors. The aforementioned characteristics impose serious requirements on the Grid programming and runtime systems if they wish to deliver efficient performance to scientific and commercial applications. To overcome the above challenges, the Grid programming and runtime systems must become autonomic or self-managing in accordance with the high-level behavior specified by system administrators. Autonomic systems are inspired by biological systems that deal with similar challenges of complexity, dynamism, heterogeneity, and uncertainty. To this end, we propose a comprehensive taxonomy that characterizes and classifies different software components and highlevel methods that are required for autonomic management of applications in Grids. We also survey several representative Grid computing systems that have been developed by various leading research groups in the academia and industry. The taxonomy not only highlights the similarities and differences of state-of-the-art technologies utilized in autonomic application management from the perspective of Grid computing, but also identifies the areas that require further research initiatives. We believe that this taxonomy and its mapping to relevant systems would be highly useful for academic-and industry-based researchers, who are engaged in the design of Autonomic Grid and more recently, Cloud computing systems.

show abstract

Managing the Execution of Large Scale MPI Applications on Computational Grids

Cited by 12 publications

References 12 publications

Distributed and dynamic self‐scheduling of parallel MPI Grid applications

Distributed and dynamic self‐scheduling of parallel MPI Grid applications

Low Cost Self-healing in MPI Applications

A taxonomy and survey on autonomic management of applications in grid computing environments

Contact Info

Product

Resources

About