Integer programming based heterogeneous CPU–GPU cluster schedulers for SLURM resource manager

Soner, Seren; Özturan, Can

doi:10.1016/j.jcss.2014.06.011

Cited by 8 publications

(16 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The new integer formulation enables us to generate higher number of bids for each job and obtain solution in a shorter period of time. When compared with the results of our previous AUCSCHED1 implementation , we see improvements in utilization, mainly as a result of being able to generate higher number of bids for each job.…”

Section: Discussionmentioning

confidence: 86%

“…In our previous work , we have used IP approach to develop two scheduling plugins for Slurm: IPSCHED and AUCSCHED1. In these plugins, the general mechanism is the same: we take a window of jobs from the front of the queue, retrieve the available resource information in the system, and then solve an IP problem.…”

Section: Related Workmentioning

confidence: 99%

“…In this work, we propose a new scheduling plug‐in for SLURM, named AUCSCHED2, which builds on our previous AUCSCHED plug‐in . AUCSCHED2 contributes two major enhancements: the first is the extension of Slurm to support generic resource moldability by specification of resource ranges.…”

Section: Related Workmentioning

confidence: 99%

“…Slurm defines accelerators that are associated with a specific node as generic resources, which can be allocated to jobs. In our previous work, we developed and released AUCSCHED1 , an auction‐based heterogeneous cluster scheduler plug‐in for Slurm, which generates and solves an integer programming (IP) problem at every job scheduling decision using CPLEX (IBM, Armonk, NY, USA) . In this work, we contribute a new auction‐based scheduler that (i) employs IP formulation with drastically reduced number of variables and (ii) offers generic resource moldability by letting users specify a range of generic resources for each job instead of a static specification at submission time.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A new auction‐based scheduler for heterogeneous systems with moldable generic resources support

Soner

Özturan

2015

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

SUMMARYSlurm resource management system is used on many TOP500 supercomputers. We present a new auctionbased heterogeneous cluster scheduler plug-in called AUCSCHED2. AUCSCHED2 contributes two major enhancements: the first is the extension of Slurm to support generic resource moldability by specification of resource ranges. The generic resources include accelerators like graphics processing unit or Xeon Phi. The current version of Slurm supports specification of node ranges but not of generic resource ranges. Such a feature can be very useful to run-time auto-tuning applications and systems that can make use of variable number of generic resources. The second enhancement involves the implementation of a new integer programming formulation in AUCSCHED2 that drastically reduces the number of variables. This allows faster solution and larger number of bids to be generated. Slurm emulation results are presented for the heterogeneous 1408 node Tsubame supercomputer, which has 12 cores and three graphics processing units on each of its nodes. AUCSCHED2 is available at https://github.com/aucsched/aucsched2.

show abstract

Section: Discussionmentioning

confidence: 86%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A new auction‐based scheduler for heterogeneous systems with moldable generic resources support

Soner

Özturan

2015

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

show abstract

“…Some of the works have explored using the RMS like TORQUE [11], SLURM [12] and HTCondor [13]. In such workaround solutions, jobs can specify their request for the GPUs but it is left up to the user to ensure that the job is executed properly on the GPU [12]. Jobs are tagged to indicate GPU and non-GPU jobs and RMSs scheduling policies have extended to be aware of GPU nodes and the respective jobs are scheduled from the job queue.…”

Section: 2using Resource Management Systemsmentioning

confidence: 99%

GPU PaaS Computation Model in Aneka Cloud Computing Environments

Ilager¹,

Wankar²,

Kune³

et al. 2019

Smart Data

View full text Add to dashboard Cite

Due to the surge in the volume of data generated and rapid advancement in Artificial Intelligence (AI) techniques like machine learning and deep learning, the existing traditional computing models have become inadequate to process an enormous volume of data and the complex application logic for extracting intrinsic information. Computing accelerators such as Graphics processing units (GPUs) have become de facto SIMD computing system for many big data and machine learning applications. On the other hand, the traditional computing model has gradually switched from conventional ownership-based computing to subscription-based cloud computing model. However, the lack of programming models and frameworks to develop cloud-native applications in a seamless manner to utilize both CPU and GPU resources in the cloud has become a bottleneck for rapid application development.To support this application demand for simultaneous heterogeneous resource usage, programming models and new frameworks are needed to manage the underlying resources effectively. Aneka is emerged as a popular PaaS computing model for the development of Cloud applications using multiple programming models like Thread, Task, and MapReduce in a single container .NET platform. Since, Aneka addresses MIMD application development 2 that uses CPU based resources and GPU programming like CUDA is designed for SIMD application development, here, the chapter discusses GPU PaaS computing model for Aneka Clouds for rapid cloud application development for .NET platforms. The popular opensource GPU libraries are utilized and integrated it into the existing Aneka task programming model.The scheduling policies are extended that automatically identify GPU machines and schedule respective tasks accordingly. A case study on image processing is discussed to demonstrate the system, which has been built using PaaS Aneka SDKs and CUDA library.

show abstract

Optimization of Data Assignment for Parallel Processing in a Hybrid Heterogeneous Environment Using Integer Linear Programming

Boiński

Czarnul

2021

The Computer Journal

View full text Add to dashboard Cite

In the paper we investigate a practical approach to application of integer linear programming for optimization of data assignment to compute units in a multi-level heterogeneous environment with various compute devices, including CPUs, GPUs and Intel Xeon Phis. The model considers an application that processes a large number of data chunks in parallel on various compute units and takes into account computations, communication including bandwidths and latencies, partitioning, merging, initialization, overhead for computational kernel launch and cleanup. We show that theoretical results from our model are close to real results as differences do not exceed 5% for larger data sizes, with up to 16.7% for smaller data sizes. For an exemplary workload based on solving systems of equations of various sizes with various compute-to-communication ratios we demonstrate that using an integer linear programming solver (lp_solve) with timeouts allows to obtain significantly better total (solver+application) run times than runs without timeouts, also significantly better than arbitrary chosen ones. We show that OpenCL 1.2’s device fission allows to obtain better performance in heterogeneous CPU+GPU environments compared to the GPU-only and the default CPU+GPU configuration, where a whole device is assigned for computations leaving no resources for GPU management.

show abstract

Integer programming based heterogeneous CPU–GPU cluster schedulers for SLURM resource manager

Cited by 8 publications

References 6 publications

A new auction‐based scheduler for heterogeneous systems with moldable generic resources support

A new auction‐based scheduler for heterogeneous systems with moldable generic resources support

GPU PaaS Computation Model in Aneka Cloud Computing Environments

Optimization of Data Assignment for Parallel Processing in a Hybrid Heterogeneous Environment Using Integer Linear Programming

Contact Info

Product

Resources

About