A Study of Adaptive Workflow Scheduling based on Machine Learning and an Extensible Simulation Environment

• Main Authors: TSAI,MENG-HAN, 蔡孟翰
• Other Authors: HUANG, KUO-CHAN
• Format: Others
• Language: en_US
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/wn2xfm

碩士 === 國立臺中教育大學 === 資訊工程學系 === 105 === With the advancement of technology and emergence of grid and cloud computing, now many large-scale scientific and engineering applications are usually constructed as workflows due to large amounts of interrelated computation and communication. Scheduling algorithms are crucial to efficient workflow execution and have become an important research topic. In this thesis, we study list-based workflow scheduling algorithms. The thesis consists of three major parts. In the rest part, we propose a new list-based workflow scheduling algorithm which can outperform current state-of-art algorithm. The second part presents a Parallel Extensible Workload Scheduling Simulator (Pewss) which we have developed to aid research work in parallel job scheduling. Based on Pewss, we conducted various simulation experiments for workflow scheduling and found that no single workflow scheduling algorithm can always achieve the best performance in all workload and platform conditions. The experimental results motivated our research work on the third part which aims to develop an adaptive workflow scheduling algorithm based on machine learning technology. The adaptive workflow scheduling algorithm is expected to achieve consistently better performance under various circumstances than any single existing workflow scheduling approach. A series of simulation experiments have been conducted to evaluate the proposed workflow scheduling algorithms. The experimental results indicate that our workflow scheduling algorithms can outperform previous scheduling methods significantly.