Simulation Modeling of Virtual Machine Based Services in Multi-Cloud Shared Access Centers

DOI

10.2991/itids-19.2019.39How to use a DOI?

Keywords

simulation modeling, Azure, DaaS, statistical analysis

Abstract

This paper addresses the problem of performance improvement of multi-cloud centers for shared access to virtual resources. Statistical study of Azure workload data was performed in Jupyter Notebooks using SciPy, StatsModels and Facebook Prophet libraries. It should be noted relatively high underutilization of the virtual resources of the Azure cloud platform. As a result of experiments the best way to simulate end-user request stream is to use the empirical cumulative distribution functions and probability mass functions calculated from the sample data. To produce high quality forecasts for time series data Prophet framework for Python and SARIMAX prediction model of StatsModels library were used. The model based on Prophet framework showed the best prediction accuracy by the Mean Absolute Error metric. Prophet framework can be used to forecast number of created and deleted virtual machines in ‘Interactive mode’, which in turn could be used in the implementation of the multi-cloud shared access center simulator.

Copyright

© 2019, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Petr Polezhaev
AU  - Leonid Legashev
AU  - Alexander Shukhman
AU  - Irina Bolodurina
PY  - 2019/05
DA  - 2019/05
TI  - Simulation Modeling of Virtual Machine Based Services in Multi-Cloud Shared Access Centers
BT  - Proceedings of the 7th Scientific Conference on Information Technologies for Intelligent Decision Making Support (ITIDS 2019)
PB  - Atlantis Press
SP  - 219
EP  - 224
SN  - 1951-6851
UR  - https://doi.org/10.2991/itids-19.2019.39
DO  - 10.2991/itids-19.2019.39
ID  - Polezhaev2019/05
ER  -