MigPF: Towards on self-organizing process rescheduling of Bulk-Synchronous Parallel applications

Righi, Rodrigo da Rosa; Gomes, Roberto de Quadros; Rodrigues, Vinicius Facco; Costa, Cristiano André da; Alberti, Antônio Marcos; Pilla, Laércio Lima; Navaux, Philippe O. A.

doi:10.1016/j.future.2016.05.004

Cited by 5 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…centralized or distributed) and are commonly implemented as sender-initiated (or push-based ) schemes. Periodic LBs target a different application set, when compared to WSs, such as Bulk Synchronous Parallel (BSP) [13], iterative, and other data-parallel applications. This is especially efficient when using persistence-based techniques, which allows the scheduler to more accurately predict the load of the tasks it is moving [].…”

Section: Related Workmentioning

confidence: 99%

PackStealLB: A scalable distributed load balancer based on work stealing and workload discretization

Freitas

Pilla

Santana

et al. 2021

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

The scalability of high-performance, parallel iterative applications is directly affected by how well they use the available computing resources. These applications are subject to load imbalance due to the nature and dynamics of their computations. It is common that high performance systems employ periodic load balancing to tackle this issue. Dynamic load balancing algorithms redistribute the application's workload using heuristics to circumvent the NP-hard complexity of the problem However, scheduling heuristics must be fast to avoid hindering application performance when distributing the workload on large and distributed environments. In this work, we present a technique for low overhead, high quality scheduling decisions for parallel iterative applications. The technique relies on combined application workload information paired with distributed scheduling algorithms. An initial distributed step among scheduling agents group application tasks in packs of similar load to minimize messages among them. This information is used by our scheduling algorithm, Pack-StealLB, for its distributed-memory work stealing heuristic. Experimental results showed that PackStealLB is able to improve the performance of a molecular dynamics benchmark by up to 41%, outperforming other scheduling algorithms in most scenarios over almost one thousand cores.

show abstract

Section: Related Workmentioning

confidence: 99%

PackStealLB: A scalable distributed load balancer based on work stealing and workload discretization

Freitas

Pilla

Santana

et al. 2021

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

show abstract

“…The quasi-optimal distributions of tasks can be obtained using the greedy strategy to distribute the work on demand, i.e. to apply dynamic load balancing techniques such as work-stealing [16], self-organising process rescheduling [31].…”

Section: The Related Workmentioning

confidence: 99%

A Three-Level Parallelisation Scheme and Application to the Nelder-Mead Algorithm

Kriauzienė

Bugajev

Čiegis

2020

Mathematical Modelling and Analysis

View full text Add to dashboard Cite

We consider a three-level parallelisation scheme. The second and third levels define a classical two-level parallelisation scheme and some load balancing algorithm is used to distribute tasks among processes. It is well-known that for many applications the efficiency of parallel algorithms of these two levels starts to drop down after some critical parallelisation degree is reached. This weakness of the twolevel template is addressed by introduction of one additional parallelisation level. s an alternative to the basic solver some new or modified algorithms are considered on this level. The idea of the proposed methodology is to increase the parallelisation degree by using possibly less efficient algorithms in comparison with the basic solver. As an example we investigate two modified Nelder-Mead methods. For the selected application, a Schro¨dinger equation is solved numerically on the second level, and on the third level the parallel Wang’s algorithm is used to solve systems of linear equations with tridiagonal matrices. A greedy workload balancing heuristic is proposed, which is oriented to the case of a large number of available processors. The complexity estimates of the computational tasks are model-based, i.e. they use empirical computational data.

show abstract

“…In the third paper, Rodrigo da Rosa Righi et al are proposing MigPF, a migration model for bulk-synchronous parallel programs on heterogeneous environments [6]. Based on the simple ideas from the TCP protocol, MigPF adapts the interval between migration calls.…”

Section: This Special Issuementioning

confidence: 99%

HPS-HDS: High Performance Scheduling for Heterogeneous Distributed Systems

Pop

Iosup

Prodan

2018

Future Generation Computer Systems

View full text Add to dashboard Cite

Heterogeneous Distributed Systems (HDS) are often characterized by a variety of resources that may or may not be coupled with specific platforms or environments. Such type of systems are Cluster Computing, Grid Computing, Peer-to-Peer Computing, Cloud Computing and Ubiquitous Computing all involving elements of heterogeneity, having a large variety of tools and software to manage them. As computing and data storage needs grow exponentially in HDS, increasing the size of data centers brings important diseconomies of scale. In this context, major solutions for scalability, mobility, reliability, fault tolerance and security are required to achieve high performance. More, HDS are highly dynamic in its structure, because the user requests must be respected as an agreement rule (SLA) and ensure QoS, so new algorithm for events and tasks scheduling and new methods for resource management should be designed to increase the performance of such systems. In this special issues, the accepted papers address the advance on scheduling algorithms, energy-aware models, self-organizing resource management, dataaware service allocation, Big Data management and processing, performance analysis and optimization.

show abstract

MigPF: Towards on self-organizing process rescheduling of Bulk-Synchronous Parallel applications

Cited by 5 publications

References 31 publications

PackStealLB: A scalable distributed load balancer based on work stealing and workload discretization

PackStealLB: A scalable distributed load balancer based on work stealing and workload discretization

A Three-Level Parallelisation Scheme and Application to the Nelder-Mead Algorithm

HPS-HDS: High Performance Scheduling for Heterogeneous Distributed Systems

Contact Info

Product

Resources

About