Multi-objective scheduling of many tasks in cloud platforms
详细信息 查看全文
- 作者：Fan Zhanga ; b ; f-zhang@tsinghua.edu.cn" class="auth_mail ; f_zhang@mit.edu" class="auth_mailAuthor Vitae ; Junwei Caob ; jcao@tsinghua.edu.cn" class="auth_mailAuthor Vitae ; Keqin Lic ; lik@newpaltz.edu" class="auth_mailAuthor Vitae ; Samee U. Khand ; samee.khan@ndsu.edu" class="auth_mailAuthor Vitae ; Kai HwangeAuthor Vitae
- 关键词：Cloud computing ; Many-task computing ; Ordinal optimization ; Performance evaluation ; Virtual machines ; Workflow scheduling
- 刊名：Future Generation Computer Systems
- 出版年：July 2014
- 年：2014
- 卷：37
- 期：Complete
- 页码：309-320
- 全文大小：2167 K
文摘
- class="listitem" id="list_l000005">
- class="label">•
We propose an ordinal optimized method for multi-objective many-task scheduling.
- class="label">•
We prove the suboptimality of the proposed method through mathematical analysis.
- class="label">•
Our method significantly reduces scheduling overhead by introducing a rough model.
- class="label">•
Our method delivers a set of semi-optimal good-enough scheduling solutions.
- class="label">•
We demonstrate the effectiveness of the method on a real-life workload benchmark.