Heterogeneous platform oriented energy consumption optimization scheduling method

Abstract

The present invention discloses a heterogeneous platform oriented energy consumption optimization scheduling method. The method comprises the following steps of: acquiring resource information of all heterogeneous physical nodes and sending the resource information to a virtual resource configuration decider; the virtual resource configuration decider calculating a virtual resource share corresponding to each heterogeneous physical node and performing modification correspondingly; refreshing configuration information of all the heterogeneous physical nodes to update the virtual resource share of each heterogeneous physical node, restarting the platform, receiving a job submitted by a user and putting the job into a job queue; a cluster platform extracting one job from the job queue and cutting the job into a plurality of tasks according to the size of a data block; and calculating the virtual resource share required by the job according to a DRF(Dominant Resource Fairness) algorithm, and the cluster platform distributing the virtual resource share required by the job to the job according to an energy consumption optimization scheduling algorithm. The heterogeneous platform oriented energy consumption optimization scheduling method provided by the present invention has the advantages of improving the platform resource scheduling efficiency by fully using a resource heterogeneity feature of the cluster platform, reducing energy consumption of platform task scheduling and the like.

Description

technical field [0001] The invention relates to an energy consumption optimization scheduling method for a distributed system, in particular to a heterogeneous platform-oriented energy consumption optimization scheduling method. Background technique [0002] Cloud computing has become a research hotspot in academia and industry. More and more universities, research institutes, IT companies and Internet companies have begun to conduct in-depth research and carry out cloud platform projects. In the cloud computing environment, the scale of distributed storage is huge, and the energy consumption is also very high. And in order to make the equipment work normally, the energy consumption also includes the energy consumption of the refrigeration equipment. In the 24×7 uninterrupted operation mode, energy consumption becomes an important part of the storage cost of the data center. In the United States, data center energy consumption doubled between 2000 and 2005. The researcher...

AI Technical Summary

Problems solved by technology

In addition, the current three mainstream scheduling algorithms on the Hadoop platform: FIFO algorithm, fair scheduling algorithm, and computing power algorithm, mainly consider the performance improvement of task scheduling in a homogeneous environment, and do not consider energy consumption optimization, especially in heterogeneous environments. Execution efficiency and energy consumption optimization problems in the environment

[0004] Although a lot of research work has been carried out in the energy consumption optimization resource scheduling direction of cloud computing in recent years, most of the algorithms do not take advantage of platform heterogeneity to better allocate resources and schedule tasks, and are based on intelligent algorithms such as machine learning algorithms. To optimize the scheduling problem is often complex and impractical

Images