Cloud computing platforms and rendering farms are driven by Internet computer graphic applications. The priority-based scheduling method is an efficient way to assign computing resources to a cloud computing platform by calculating the ratio of resources based on payments in advance. Typically, the job arrival rates of users on a cloud computing platform are not evenly distributed. Thus, low-paying users can grab more resources by increasing their job arrival rates. The purpose of this paper is to propose a new method that adjusts the resource allocation ratio based on the level of payment within a specified time interval. A system administrator can determine a user’s resource quota’s validity by controlling the length of the time interval. According to our experimental results, our new method effectively allocates resources based on payment level and reduces the impact of different job arrival rates.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

A Job Scheduling Method for Ensuring Long-Term Workload Fairness in Cloud Platforms

  • Lung-Pin Chen,
  • Leu-Fang Yei,
  • Hsin-Ta Chiao

摘要

Cloud computing platforms and rendering farms are driven by Internet computer graphic applications. The priority-based scheduling method is an efficient way to assign computing resources to a cloud computing platform by calculating the ratio of resources based on payments in advance. Typically, the job arrival rates of users on a cloud computing platform are not evenly distributed. Thus, low-paying users can grab more resources by increasing their job arrival rates. The purpose of this paper is to propose a new method that adjusts the resource allocation ratio based on the level of payment within a specified time interval. A system administrator can determine a user’s resource quota’s validity by controlling the length of the time interval. According to our experimental results, our new method effectively allocates resources based on payment level and reduces the impact of different job arrival rates.