Performance Tuning for Big Data Applications in Docker Containers

	Performance Tuning for Big Data Applications in Docker Containers
	Kejiang Ye; Yunjie Ji
	2017
会议日期	2017
会议地点	Shenzhen
英文摘要	Docker container technology is experiencing a rapidly development with the support from industry and being widely used in large scale production cloud environment. Speedy launching time and tiny memory footprint are two main benefits of container virtualization. But the performance of Spark big data applications running on Docker containers is still not clear enough due to the complex parameter configuration and interference between neighbor containers. This paper investigates the impacts of docker configuration and resource interference on application performance in a typical container virtualized computing environment. In particular, we first conduct a series of experiments to measure the performance impact by adjusting the docker configuration parameters, such as resource limits, and observe the Spark performance is not linear with increasing resource allocation for containers. Then, we evaluate the interference between multiple containers by controlling the resource competition and identify the performance interference phenomenon between multiple containers. Finally, we propose a performance prediction model based on the Support Vector Regression to predict the docker application performance with different configurations and resource competition settings. Experimental results show the prediction error is less than 10% for all the four typical Spark applications.
语种	英语
内容类型	会议论文
源URL	[http://ir.siat.ac.cn:8080/handle/172644/12652]
专题	深圳先进技术研究院_数字所
作者单位	2017
推荐引用方式 GB/T 7714	Kejiang Ye,Yunjie Ji. Performance Tuning for Big Data Applications in Docker Containers[C]. 见:. Shenzhen. 2017.