Virtualization and computing clouds technology application within complex distributed simulated systems creation

Mathematica modeling, numerical technique and program complexes


Аuthors

Kondrashin M. A.*, Arsenov O. Y.**, Kozlov I. V.***

Central Research Institute of Machine Building, 4, Pionerskaya st., Korolev, Moscow region, 141070, Russia

*e-mail: mak@mcc.rsa.ru
**e-mail: aoy@mcc.rsa.ru
***e-mail: kozloviv@mcc.rsa.ru

Abstract

This article considers the approaches to developing complex distributed simulation systems for conducting experiments on test subject using up-to-date methods and means of virtualization and cloud computation organization.

Simulation means modeling performed with computers. It allows automatization of experiments conducted on a tested system. Distributed modeling is the modeling employing multiprocessor computers and/or computer network. The main requirement, defining the necessity of distributed modeling architecture implementation is complex simulation systems integration. The main advantages of the distributed modeling systems are as follows: the possibility of multiple stand-alone hardware sets integration while conducting an experiment; a new sample development verification by one of the distributed system element substituting; developing simulators for teaching personnel, operating the real system. Labor consuming process of preparing for conducting an experiment and original setting of the interconnections between the system elements; requirement of the exact notion of necessary computing resources at the stage of modeling system design; the need for implementation of special algorithms for hardware computing resources balancing can be referred to as the drawbacks of distributed simulation systems while developing complex simulation systems.

Today, in the age of computer technologies we suggest using the virtualization and computing clouds technology when developing distributed modeling complexes. Such technology helps you to get rid of the most of the disadvantages when building the distributed systems for modeling. Virtualization is a hardware-software opportunity of computer/server physical resources (processor, memory, storage and network adapter) utilization optimization through their representation in the shared access between units (virtual machines) logically isolated from each other.

The paper suggests an approach for the distributed modeling systems developing, using the virtualization and computing clouds technologies. We provide the description of the technology and its advantages, foremost among which is an easily scalable distributed modeling systems creation.

The simulation complex of the SC GCC elements and SC GСС in general under development at TsNIIMash (Central scientific research institute of mechanical engineering) is described.

The paper shows that the virtualization and cloud computation methods utilization when building the distributed modeling systems allow get rid of the most of the problems emerging from building analogous systems in a traditional way.A description of the developed applications allowing automatize the virtualization platform control process is also presented.

Keywords:

simulation, distributed system, virtualization, computing clouds

References

  1. Karpov Ju.G. Imitatsionnoe modelirovanie sistem (Systems Simulation), St. Petersburg, BHV-Peterburg, 2005, 400 p.

  2. Kostin P.S., Vereshhagin Ju.O., Voloshin V.A. Trudy MAI, 2015, no. 81: http://www.mai.ru/science/trudy/published.php?ID=57735

  3. Lohmatkin V.V. Trudy MAI, 2015, no. 81: http://www.mai.ru/science/trudy/published.php?ID=57787

  4. Samojlenko A. Virtualizatsiya serverov v srednem i malom biznese, URL: http://www.ixbt.com/cm/virtualization-smb.shtml

  5. Kazak M. Virtualizatsiya: tekhnologiya, kotoraya izmenit IT-mir, URL: http://www.infosecurity.ru/_gazeta/content/080627/article01.shtml#top


Download

mai.ru — informational site MAI

Copyright © 2000-2019 by MAI

Вход