TRAP@NCI

Docker Container Reactive Scalability and Prediction of CPU Utilization Based on Proactive Modelling

Shanmugam, Aravind Samy (2017) Docker Container Reactive Scalability and Prediction of CPU Utilization Based on Proactive Modelling. Masters thesis, Dublin, National College of Ireland.

[img]
Preview
PDF (Master of Science)
Download (5MB) | Preview

Abstract

The resource utilization of cloud-based applications has been used for capacity planning and forecasting resource demand. Cloud scalability provides significant scaling techniques to identify real-time demand as well the future demand to reduce resource wastage and network bandwidth. The project focuses to tackle the Docker container reactive scalability and prediction of CPU utilization using pro-active modelling. The docker container reactive scaling is structured based on threshold policies on cloud and candidate performed scaling in and scaling out of docker container web services built on CPU utilization threshold. The CPU metrics is perceived on cadvisor dashboard where HAproxy is utilized to distribute load to container web services based on round-robin algorithm. In addition, proactive scaling is performed using ARIMA time series analysis model. The public cloud Amazon Web Services is used to perform container reactive scaling and containerized open source jupyter notebook application is utilized to perform prediction. Based on CPU utilization data observed on amazon cloud, the proactive ARIMA model is trained. The parameter combination values are evaluated using Akaike Information Criterion value to get best results for forecasting CPU utilization. The paper explores the evaluation results of container reactive scaling and ARIMA prediction CPU utilization. In addition, the state of the art results are also demonstrated.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science

T Technology > T Technology (General) > Information Technology > Cloud computing
Divisions: School of Computing > Master of Science in Cloud Computing
Depositing User: CAOIMHE NI MHAICIN
Date Deposited: 22 Nov 2017 09:41
Last Modified: 22 Nov 2017 09:41
URI: http://trap.ncirl.ie/id/eprint/2884

Actions (login required)

View Item View Item