Malleable Task-Graph Scheduling with a Practical Speed-Up Model

Marchal, Loris; Simon, Bertrand; Sinnen, Oliver; Vivien, Frédéric

doi:10.1109/tpds.2018.2793886

Cited by 9 publications

(5 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They used a three-step algorithm to solve the problem: parallel degree allotment, task scheduling and frequency assignment. Marchal et al [29] proposed a practical speedup model for graphs of MPTs, which takes trade-off between tractability and accuracy. They proposed model-optimized variants of the existing algorithms PROPMAPPING and FLOWFLEX, and proved them are 2-approximation algorithms.…”

Section: Related Workmentioning

confidence: 99%

“…In this section, first, we introduce the MPT model that supports time-slice. Here, we give an example of MPTs-WRF (Weather Research and Forecasting Model) [29]. WRF is widely used in the weather forecast.…”

Section: Moldable Parallel Task Supporting Time-slice Modelmentioning

confidence: 99%

“…The reason is that the system needs more time to exchange data between different resources when the MPT has a large parallelism (than a constant). Like most of parallel tasks, WRF [29] has a nonlinear speedup.…”

Section: Moldable Parallel Task Supporting Time-slice Modelmentioning

confidence: 99%

See 2 more Smart Citations

A Scheduling Method of Moldable Parallel Tasks Considering Speedup and System Load on the Cloud

Zhang

2019

IEEE Access

View full text Add to dashboard Cite

The moldable parallel task (MPT) is a kind of parallel task that their sub-tasks hold the resources exclusively, which has been widely used in different areas. Our paper focuses on the scheduling of moldable tasks when every sub-task supports time-slice. The time-slice is a consecutive time that the sub-task holds the resources exclusively. After every time-slice, the sub-task can be canceled, suspended, or continued. We give the model of MPTs and propose a scheduling method for MPTs: MC (a heuristic scheduling method supporting time-slice model on the cloud). The simulation results show that, even under a forecast accuracy of system load under 90% and 95%, MC reduces average waiting time and average execution time; at the same time, MC has a lower value in the percentages of unfinished tasks. INDEX TERMS Moldable parallel task, speedup, parallelism, scheduling method.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Moldable Parallel Task Supporting Time-slice Modelmentioning

confidence: 99%

See 1 more Smart Citation

A Scheduling Method of Moldable Parallel Tasks Considering Speedup and System Load on the Cloud

Zhang

2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…We will refer to this in the remainder of this paper as dynamicity, which has been shown to improve overall system efficiency with respect to various optimization goals, such as throughput, energy or job queue latency [13,19,20,23]. Dynamicity can also be advantageous for individual applications: Avoiding a-priori resource fixation and using the global scheduler perspective can improve individual resource efficiency and support urgent computing [6,15,16,20].…”

Section: Introductionmentioning

confidence: 99%

Towards Dynamic Resource Management with MPI Sessions and PMIx

Huber

Streubel

Comprés

et al. 2022

Proceedings of the 29th European MPI Users' Group Meeting

View full text Add to dashboard Cite

Job management software on peta-and exascale supercomputers continues to provide static resource allocations, from a program's start until its end. Dynamic resource allocation and management is a research direction that has the potential to improve the efficiency of HPC systems and applications by dynamically adapting the resources of an application during its runtime. Resources can be adapted based on past, current or even future system conditions and matching optimization targets. However, the implementation of dynamic resource management is challenging as it requires support across many layers of the software stack, including the programming model.In this paper, we focus on the latter and present our approach to extend MPI Sessions to support dynamic resource allocations within MPI applications. While some forms of dynamicity already exist in MPI, it is currently limited by requiring global synchronization, being application or application-domain specific, or by suffering from limited support in current HPC system software stacks.We overcome these limitations with a simple, yet powerful abstraction: resources as process sets, and changes of resources as set operations leading to a graph-based perspective on resource changes. As the main contribution of this work, we provide an implementation of this approach based on MPI Sessions and PMIx. In addition, an illustration of its usage is provided, as well as a discussion about the required extensions of the PMIx standard. We report results based on a prototype implementation with Open MPI using a synthetic application, as well as a PDE solver benchmark on up to four nodes and a total of 112 cores. Overall, our results show the feasibility of our approach, which has only very moderate overheads. We see this first proof-of-concept as an important step towards resource adaptivity based on MPI Sessions.

show abstract

“…That work either supposes that the task has a forecast speedup under different parallelisms [19], or supposes that we know the details of the DAG of the parallel task. However, it is very difficult to get an accurate forecast of the speedup under different parallelisms [20].…”

Section: Introductionmentioning

confidence: 99%

Resource management for moldable parallel tasks supporting slot time in the Cloud

2019

KSII TIIS

View full text Add to dashboard Cite

Moldable parallel tasks are widely used in different areas, such as weather forecast, biocomputing, mechanical calculation, and so on. Considering the deadline and the speedup, scheduling moldable parallel tasks becomes a difficulty. Past work majorly focuses on the LA (List Algorithms) or OMA (Optimizing the Middle Algorithms). Different from prior work, our work normalizes execution time and makes all tasks have the same scope in normalized execution time: [0,1], and then according to the normalized execution time, a method is used to search for the reference execution time without considering the deadline of tasks. According to the reference execution time, we get an initial scheduling result based on AFCFS (Adaptive First Comes First Served) policy. Finally, a heuristic approach is used to improve the performance of the initial scheduling result. We call our method HSRET (a Heuristic Scheduling method based on Reference Execution Time). Comparisons to other methods show that HSRET has good performance in AWT (Average Waiting Time), AET (Average Execution Time), and PUT (Percentages of Unfinished Tasks).

show abstract

Malleable Task-Graph Scheduling with a Practical Speed-Up Model

Cited by 9 publications

References 44 publications

A Scheduling Method of Moldable Parallel Tasks Considering Speedup and System Load on the Cloud

A Scheduling Method of Moldable Parallel Tasks Considering Speedup and System Load on the Cloud

Towards Dynamic Resource Management with MPI Sessions and PMIx

Resource management for moldable parallel tasks supporting slot time in the Cloud

Contact Info

Product

Resources

About