Performance comparison of jobs with and without task splitting
碩士 === 國立臺灣科技大學 === 資訊管理系 === 88 === An important part of a distribution/parallel system design is the workload sharing among the processors. This includes partitioning the arriving jobs into tasks that can be executed in parallel, assigning the tasks to processors and scheduling the task execution...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2000
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/31368840306641028019 |
| Summary: | 碩士 === 國立臺灣科技大學 === 資訊管理系 === 88 === An important part of a distribution/parallel system design is the workload sharing among the processors. This includes partitioning the arriving jobs into tasks that can be executed in parallel, assigning the tasks to processors and scheduling the task execution on each processor.
In Anthony karageorgos and Helen karatza’s (1997) paper they examine the efficiency of two task routing strategies and three nonpreemptive task scheduling policies in conjunction with job resequencing before departure. The job must depature in the order of their arrival.
We analyze by simulation the difference between Fork-Join System and single processor model. Single processor model examines the efficiency in the mean response time and system throughput of one task routing and three task scheduling strategies. For the performance measures of mean response time and throughput, empirical results show that single processor model outperforms the Fork-Join System.
|
|---|