International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May-2014 110
ISSN 2229-5518
Comparative Study of Simulation Tools in Cloud
Computing Environment
Ranu pandey, Sandeep Gonnade
Abstract— Cloud computing includes delivery of dependable, secure, fault-tolerant tolerant, maintainable, and scalable infrastructures for facilitating web based application services. These applications have diverse composition, setup, and deployment requirements. As the adoption and deployment of cloud computing increases, it is discriminating to assess the performance of cloud computing environments. Modeling and simulation are suitable for assessing performance and security issues. Cloud simulators are needed for cloud framework testing to abatement the intricacy and separate out quality concerns. A few cloud simulators have been particularly created for performance testing of cloud computing environments. In this paper we study diverse sort of cloud simulators .Thereafter, we rundown the examinations of distinctive tools by some criteria.
Index Terms—. Cloud Computing, Cloud Simulation tool, CloudSim, CloudAnalyst, EMUSIM, GreenCloud, MDCSim, NetworkCloud, ican cloud.
Abbreviations—Automatic Emulation Framework (AEF); Data Center (DC); Network Simulator 2 (NS2); Virtual Machine (VM).
—————————— ——————————
Cloud computing is a region that is encountering a quick development both in the academia and industry. This technol- ogy, which points at offering appropriated, virtualized, and adaptable assets as utilities to end clients, can possibly support full acknowledgment of "computing as a utility" .Along with the development of the cloud technology, new conceivable outcomes for internet-based application development are ris- ing. These new application models could be grouped into two catagories: on one side, there are the cloud service providers that are eager to give substantial scale computing infrastruc- ture at a cost based fundamentally with respect to use designs. It eliminates the beginning high-take for application develop- ers of environment set up application deployment. On the other side there are large scale programming frameworks suppliers, which create applications, for example, social net- working and e-commerce, which are picking up fame on the Internet. These applications can profit enormously from cloud infrastructure administrations to minimize expenses and en- hance administration quality to end clients.
Previously, development of such applications required obtain-
ing of servers with a fixed capacity fit to handle the normal application peak demand, establishment of the entire pro- gramming framework of the stage supporting the application, and design of the application itself.
• Sandeep Gonnade is currently serving as Asst. Professor in Computer Science
engineering in MATS University, India, PH-. 09039822582 E-mail:
sandeep_gonnade@yahoo.co.in
Anyhow the servers were underutilized more often than not on account of crest movement happens just at particular brief time periods. With the coming of the Cloud, organization and facilitating got to be less expensive and simpler with the utili- zation of pay-for every use, adaptable elastic framework ser- vices provided by cloud suppliers.
When these two ends are united, a few elements that effect the net profit of cloud might be watched. Some of these elements incorporate geographic distribution of client bases, capabilities of the internet infrastructure within those geographic areas, dynamic nature of usage patterns of the user bases, and capa- bilities of cloud services in terms of adaptation or dynamic reconfiguration, among others.
A comprehensive study of the whole problem in the real
Internet platform is extremely difficult, because it requires interaction with several computing and network elements that cannot be controlled or managed by application developers. Furthermore, network conditions cannot be predicted nor con- trolled, and it also impacts quality of strategy evaluation.
A more viable alternative is the use of cloud simulation tools. Cloud simulators are required for cloud environment testing to decrease the complexity and separate out quality concerns. They enable performance analysts to examine sys- tem behavior by focusing on quality issues of specific compo- nent under different scenarios [Xiaoying Bai et al., 2011]. These tools open up the possibility of evaluating the hypothesis in a controlled environment where one can easily reproduce re- sults. Simulation-based approaches offer significant benefits to IT companies by allowing them to test their services in repeat- able and controllable environment and experiment with dif- ferent workload mix and resource performance scenarios on
IJSER © 2014 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May-2014 111
ISSN 2229-5518
simulated infrastructures for developing and testing adaptive application provisioning techniques [Calheiros et al., 2011].
None of the current cloud system simulators offer the environment that can be directly used for modeling Cloud computing environments. But Cloud Simulators which are generalized and extensible simulation frameworks that allow seamless modeling, simulation, and experimentation of emerging Cloud computing infrastructures and application services. By using cloud simulators, researchers and develop- ers can test the performance of a newly developed application service in a controlled and easy to set-up environment. The vast features of cloud simulators would speed up the devel- opment of new application provisioning algorithms for cloud computing. This paper first gives background about various simulators available. Section 3 defines cloud simulators that are available such as CloudAnalyst, GreenCloud, Network- Cloud, EMUSIM and MDCSim. ,iCancloud,virtual cloud.In the section 4, it Compares all Cloud Simulators with respect to different criteria.
In the previous decade, Grids [foster & Kesselman, 1999] have developed as the infrastructure for delivering high- performace service for compute and data-intensive scientific applications. To support research, development and testing of new grid segments, policies and middleware, a few grid simu- lators, for example, Gridsim, Simgrid, Optor-Sim and Gang- sim have been proposed. Simgrid is a generic framework for simulation of distributed applications in grid platforms. Gang- sim is a grid simulation toolkit that provides help for model- ing of grid-based virtual organizations and assets. Then again, Gridsim is an event driven simulation tool stash for heteroge- neous grid assets. It backs modeling of grid elements, clients, machines, and network including network traffic yet none of these can support the infrastructure and application level re- quirements emerging from cloud processing ideal model. In particular, there is no support in existing grid simulation toolkits for modeling of on-demand virtualization enabled resource and application administration. Further, cloud infra- structure modeling and simulation toolkits must give backing to economic entities, for example, cloud brokers and cloud exchange for enabling ongoing exchanging of services. Among the currently available simulator talked about, just Gridsim offers support for investment driven resource management and application scheduling simulation.
Simulation these days assumes an expanding part in the as- sessment of conceivable results and circumstance analysis. It is the limitation of the operation of a real world methodology or
system (Banks 1998). Simulation modeling and analysis is the procedure of creating and exploring different avenues regard- ing a computerized mathematical model of a physical frame- work (Chang 2004). Simulation is the route how to research the mod-
el.
Figure 1: Development of simulation tools architecture
Cloud simulators give a summed up and extensible simulation framework that enables modeling, simulation, and experimen- tation of emerging cloud computing infrastructures and appli- cation .These are the following simulators that support the cloud environment:
CloudSim is a simulation application which enables seamless modeling, simulation, and experimentation of cloud compu- ting and application services [Calheiros et al., 2009; 2011; Buy- ya et al., 2009] due to the problem that existing distributed system simulators were not applicable to the cloud computing environment. Evaluating the performance of cloud provision- ing policies, services, application workload, models and re- sources performance models under varying system, user con- figurations and requirements is difficult to achieve. To over- come this challenge, CloudSim can be used. CloudSim is new, generalized and extensible simulation toolkit that enables seamless modeling, simulation and experimentation of emerg- ing cloud computing system, infrastructures and application environments for single and internetworked clouds. In simple words, CloudSim [Rahul Malhotra & Prince Jain, 2013] is a development toolkit for simulation of Cloud scenarios. CloudSim is not a framework as it does not provide a ready to use environment for execution of a complete scenario with a specific input. Instead, users of CloudSim have to develop the cloud scenario it wishes to evaluate, define the required out- put, and provide the input parameters [Dr. Rahul Malhotra & Prince Jain, 2013].
CloudSim is invented and developed as CloudBus Project at
the University of Melbourne, Australia. The CloudSim toolkit supports system and behavior modeling of cloud system components such as data centers, virtual machines (VMs) and resource provisioning policies. It implements generic applica-
IJSER © 2014 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May-2014 112
ISSN 2229-5518
tion provisioning techniques that can be extended with ease and limited efforts. CloudSim helps the researchers and de- velopers to focus on specific system design issues without get- ting concerned about the low level details related to cloud- based infrastructures and services [Wickremasinghe, 2009]. CloudSim is an open source web application that launches preconfigured machines designed to run common open source robotic tools, robotics simulator Gazebo.
Fig. 2 The CloudSim architecture
The users could analyze specific system problems through CloudSim, without considering the low level details related to Cloud based Infrastructures and services [Wei Zhao et al.,
2012].
CloudAnalyst was derived from CloudSim and extends some of its capabilities and features proposed [Wickremasinghe,
2009; Wickremasinghe & Calheiros, 2010]. CloudAnalyst sepa- rates the simulation experimentation exercise from a pro- gramming exercise. It also enables a modeler to repeatedly perform simulations and to conduct a series of simulation ex- periments with slight parameters variations in a quick and easy manner. CloudAnalyst can be applied to examining be- havior of large scaled Internet application in a cloud environ- ment.
.
. Figure 3. CloudAnalyst architecture
The main features of CloudAnalyst are the following.:
1. Easy to use Graphical User Interface (GUI). Cloudanalyst is outfitted with a easy to use graphical user interface (see Fig- ure 4) that empowers clients to set up experiments rapidly and effectively.
2. Ability to define a simulation with a high degree of con-
figurability and flexibility. Simulation of complex frame-
works, for example, Internet applications relies on numerous parameters. Commonly, values for those parameters need to be arbitrarily assumed or decided through a procedure of ex- perimentation. Cloudanalyst gives modelers a high level of control over the experiment, by displaying elements and con- figuration options, for example, Data Center, whose hardware configuration is characterized regarding physical machines made out of processors, storage devices, memory and internal band-width; Data Center virtual machine determination as far as memory, storage and bandwidth quota; Resource assign- ment policies for Data Centers (e.g., time-shared vs. space- shared); Users of the application as groups and their distribu- tion both geologically and transiently; Internet progress with configuration options for network delays and available band- width; Service Broker Policies that control which segment of aggregate client base is services by which data center at a giv- en time; and simulation span in minutes, hours or days.
3. Repeatability of experiments.
Cloudanalyst permits modelers to save simulation experi- ments input parameters and brings about the manifestation of XML records so the examinations might be repeated. The un- derlying Cloudsim simulation framework ensures that repeat- ed experiments yield indistinguishable effects.
4. Graphical output
. Cloudanalyst is capable of creating graphical yield of the
simulation results in the form of tables and graphs, which is desirable to viably outline the substantial measure of facts that is gathered throughout the simulation. Such a viable presenta- tion helps in distinguishing the important examples of the yield parameters and aides in comparisons between related parameters. In the current version of Cloudanalyst, the ac- companying factual measurements are professional produced as yield of the reproduction: Response time of the reenacted provision; general normal, least and maximum reaction time of all client solicitations reproduced; response time organized by client gatherings, spotted inside geographical locales; reac- tion time masterminded by time, demonstrating the example of progressions in requisition use throughout the day; use ex- amples of the provision; number of clients orchestrated by time or areas of the world, and the general impact of that use on the server farms facilitating the requisition; time taken by server farms to administration a client demand; general ask for genius cessing time for the whole recreation; normal, least and greatest appeal preparing time by every server farm; reac- tion time variety example throughout the day as the heap changes; and points of interest of expenses of the operation.
IJSER © 2014 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May-2014 113
ISSN 2229-5518
Figure 4. Cloud Analyst GUI
5. Use of consolidated technology and ease of extension.
CloudAnalyst is based on a modular design that can be easily extended. It is developed using the following technologies: Java (the simulator is developed 100% on Java platform, using Java SE 1.6); Java Swing (the GUI component is built using Swing components); CloudSim (CloudSim features for model- ing data centers is used in CloudAnalyst); and SimJava [6] (some features of this tool are used directly in CloudAnalyst
Network CloudSim is an extension of CloudSim as a simula- tion framework which supports generalized applications such as high performance computing applications, workflows and e-commerce [Buyya et al., 2009]. Network CloudSim uses Network Topology class which implements network layer in CloudSim, reads a BRITE file and generates a topological net- work. In network CloudSim, the topology file contains nodes, number of entities in the simulation which allows users to in- crease scale of simulation without changing the topology file. Each CloudSim entity must be mapped to one BRITE node to allow proper work of the network simulation. Each BRITE node can be mapped to only one entity at a time. Network CloudSim allows for modeling of Cloud data centers utilizing bandwidth sharing and latencies to enable scalable and fast simulations. Network CloudSim structure supports designing of the real Cloud data centers and mapping different strate- gies. Information of network CloudSim is used to simulate latency in network traffic of CloudSim. This simulation framework which supports the modelling of essential data center resources such as network and computational re- sources, and wide variety of application models such as paral- lel application, workflow and parametric sweep.
Figure 5: The CloudSim Architecture with
NetworkCloudSim elements
EMUSIM is an integrated architecture [Calheiros et al., 2012] to anticipate service’s behavior on cloud platforms to a higher standard [Calheiros et al., 2011; Wei Zhao et al., 2012]. EMUSIM combines emulation and simulation to extract in- formation automatically from the application behavior via emulation and uses this information to generate the corre- sponding simulation model. Such a simulation model isthen used to build a simulated scenario that is closer to the actual target production environment in application computing re- sources and request patterns. Information that is typically not disclosed by platform owners, such as location of virtual ma- chines and number of virtual machines per host in a given time, is not required by EMUSIM. EMUSIM is built on top of two software systems: Automated Emulation Framework (AEF) for emulation and CloudSim for simulation.
MDCSim is a commercial discrete event simulator devel- oped at the Pennsylvania State University. It helps the analyz- er to model unique hardware characteristics of different com- ponents of a data center such as servers, communication links and switches which are collected from different dealers
and allows estimation of power consumption. MDCSim is the
most prominent tool to be used as it has low simulation over- head and moreover its network package maintains a data cen- ter topology in the form of directed graph [Dr. Pawan Kumar
& Gaganjot Kaur].
GreenCloud is a Cloud Simulator that have green cloud computing approach with confidently, painlessly, and success- fully. In other words, GreenCloud is developed as an ad- vanced packet level cloud network simulator with concentra- tion on cloud communication [Kliazovich et al., 2010]. Green- Cloud extracts, aggregates and makes fine grained infor-
IJSER © 2014 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May-2014 114
ISSN 2229-5518
mation about the energy consumed by computing and com- munication elements of the data center equipment such as computing servers, network switches and communication links [Wei Zhao et al., 2012; http://www.isi.edu/nsnam/ns/] available in an unprecedented fashion. Moreover, GreenCloud offers a thorough investigation of workload distributions. In particular, a special focus is devoted to accurately capture communication patterns of currently deployed and future da- ta center architectures. GreenCloud can act as Cloud Bridge [http://gogreencloud.com]. In simple words, GreenCloud is the practice of designing, manufacturing, using and disposing computing resources with minimal environmental damage. The Green Cloud is a supercomputing project under active development at the University of Notre Dame. Green Cloud provides a virtual computing platform by using grid heating which reduces cluster upkeep costs.
Figure 6: A User View of Green Cloud
GreenCloud Simulator is an extension of Network Simula- tor (NS2) simulator [Wei Zhao et al., 2012; http://www.isi.edu/nsnam/ns/]. GreenCloud simulator im- plements a full TCP/IP protocol reference model which allows integration of different communication protocols with the simulation. The only drawback of Green Cloud Simulator is that it confines its scalability to only small data centers due to very large simulation time and high memory requirements
tions facilitating Internet services to meet clients' quality of service requirements.
2. To minimize utilization of electric power by enhancing power management, dynamically managing and configuring power aware capability of system devices [http://www.cloudbus.org/greencloud/].
3. To Provide detailed simulators.
4. To investigate energy efficiency and measure cloud perfor- mance.
Greencloud can lessen Data Center Power Consumption by:
1. Workload consolidation through DC virtualization.
2. By statistical multiplexing incentives prompting aggressive-
ly bringing down Opex.
3. By enhancing manageability by decreasing host count..
iCancloud is a simulation platform that is created by a re- search group(ARCOS) at Universidad carlos iii de madrin spain that meant to model and recreate cloud computing frameworks, which is focused to those clients who deal with those sorts of frameworks. The fundamental goal of iCancloud is to foresee the exchange offs between expense and perfor- mance of a given set of applications executed in a particular hardware, and after that give to clients helpful data about such costs. Nonetheless, iCancloud could be utilized by an extensive variety of clients, from fundamental dynamic clients to developers of large distributed applications. It gives flexible cloud hypervisor module and an amicable client GUI to sim- plify the generation and customization of huge distribut mod-
Energy and Busi- ness Efficiency
Cloud Computing
Figure7 : Pictorial View of Green Computing
GreenCloud Aim
1. To create high-end computing systems, for example, Clus-
ters, Data Centers, and Clouds that distribute assets to applica-
els.
Figure8: Configuration of ican cloud process
IJSER © 2014 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May-2014 115
ISSN 2229-5518
The number of simulation environments for cloud compu- ting data centers accessible for open utilization is constrained. The Cloudsim simulator is most likely the most modern around the simulators over-viewed. The Mdcsim simulator is a moderately new data center simulator created at the Penn- sylvania State University. It is supplied with specific hardware chracteristics of data server parts, for example, servers, com- munication links and changes from diverse vendor. Table 1 looks at different Cloudsim simulators through correlation of their qualities, for example, platform, language, networking, simulator type and availability [dzmitry Kliazovich et al.,
2010]. The proposed Greencloud simulator is created as an
extension of the Ns2 network simulator which is coded in C++ with a layer of Otcl libraries implemented on top of it. It is a packet level simulator. On the contrary, Cloudsim and
Mdcsim are event based simulators which avoid building and processing small objects exclusively. Such a strategy decreases simulation time considerably, improves scalability, however lacks in the simulation exactness. Both Greencloud and Cloudsim simulators are discharged under open source GPL license. The Mdcsim simulator is presently not available for open download which is a commercial product [dzmitry Kliazovich et al., 2010]. Condensing, short simulation times are provided by Cloudsim and Mdcsim actually for large data centers because of their event based nature, while Greencloud offers a change in the simulation precision keeping the simula- tion time at the reasonable level. None of the tools offer user- friendly graphical interface. The GreenCloud supports cloud computing workloads with deadlines, but only simple sched- uling policies for single core servers are implemented. The ican cloud simulator provides flexiblecloud hypervisor mod- ule and a friendly user GUI to ease the generation and custom- ization of large distributed models The MDCSim workloads are described with the computational requirements only and require no data to be transferred. Communication details and the level of energy models support are the key strengths of the GreenCloud which are provided via full support TCP/IP pro- tocol reference model and packet level energy models imple- mented for all data center components [Dzmitry Kliazovich et al., 2010
Table1. Comparision of various cloudsim
Table2. Comparision of various cloud simulators
Cloud computing has been one of the quickest developing parts in IT industry. Simulation based methodologies get to be popular in industry and the educated community to assess cloud computing frameworks, applications behaviors and their security. A few simulators have been particularly devel- oped for performance analysis of cloud computing environ- ments including Cloudsim, Greencloud, Networkcloudsim, Cloudanalyst, EMUSIM and Mdcsim yet the number of simu- lation environments for cloud computing data centers availa- ble for public use is restricted. The Cloudsim simulator is pre- sumably the most refined among the simulators reviewed. The Mdcsim simulator is a relatively fresh data center simulator.
IJSER © 2014 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May-2014 116
ISSN 2229-5518
Finally we provide the overview on the current development status of cloud simulators in cloud environment.
[1] I. Foster & C. Kesselman (1999), “The Grid: Blueprint for a New Com- puting Infrastructure”, Morgan Kaufmann.
[2] R.N. Calheiros, Rajiv Ranjan, César A.F. De Rose & Rajkumar Buyya (2009), “CloudSim: A Novel Frame work for Modeling and Simulation of Cloud Computing Infrastructure and Services”, Technical Report of GRIDS Laboratory, The University of Melbourne, Australia.
[3] R. Buyya, R. Ranjan & R.N. Calheiros (2009), “Modeling and Simula- tion of Scalable Cloud Computing Environments and the CloudSim Toolkit: Challenges and Opportunities”, The International Conference on High Performance Computing and Simulation, Pp. 1–11.
[4] B. Wickremasinghe (2009), “CloudAnalyst: A CloudSim based Tool for Modeling and Analysis of Large Scale Cloud Computing Environments”, MEDC Project Report.
[5] B. Wickremasinghe & R.N. Calheiros (2010), “CloudAnalyst: A CloudSim based Visual Modeler for Analyzing Cloud Computing Envi- ronments and Applications”, 24th International Conference on Advanced Information Networking and Application, Pp. 446–452. and Simulation, SCS, Jan. 2008.
[6] F. Howell and R. Macnab, “SimJava: a discrete event simulation library
for Java,” Proc. of the 1st International Conference on Web-based Model- ing.
[7] D. Kliazovich, P. Bouvry & S.U. Khan (2010), “GreenCloud: A Packet- Level Simulator of Energy Aware Cloud Computing Data Centers”, IEEE Global Telecommunications Conference, Pp. 1–5.
[8] Dzmitry Kliazovich, Pascal Bouvry & Samee Ullah Khan (2010), “GreenCloud: A Packet-Level Simulator of Energy,springe
Science+Business Media, LLC, Luxembourg.
[9] Xiaoying Bai, Muyang Li, Bin Chen, W.T. Tsai & J. Gao (2011), “Cloud Testing Tools”, Proceedings of the 6th IEEE International Symposium on Ser- vice Oriented System Engineering, Pp. 1–12.
About Corresponding Author
Corresponding Author, Ranu Pandey is pursuing M.Tech. Final Semester, De- partment of Computer science Engineering, School of Engineering and IT MATS UNIVERSITY,RAIPUR(C.G.).
[10] R.N. Calheiros, Rajiv Ranjan, Anton Beloglazov, Cesar A.F. De Rose & Rajkumar Buyya (2011), “CloudSim: A Toolkit for Modeling and Simula- tion of Cloud Computing Environments and Evaluation of Resource Pro- visioning Algorithms”, Software: Practice and Experience, Vol. 41, No. 1, Pp.
23–50.
[11] http://gogreencloud.com, Green Cloud Technologies, 2012
[12] Wei Zhao, Yong Peng, Feng Xie & Zhonghua Dai (2012), “Modeling and Simulation of Cloud Computing: A Review”, IEEE Asia Pacific Cloud Computing Congress (APCloudCC), Pp. 20–24.
[13] R.N. Calheiros, Marco A.S. Netto, César A.F. De Rose & Rajkumar Buyya (2012), “EMUSIM: An Integrated Emulation and Simulation Envi- ronment for Modeling, Evaluation, and Validation of Performance of Cloud Computing Applications”, Software: Practice and Experience, Pp.
1–18.
[14] Dr. Rahul Malhotra & Prince Jain (2013), “An EMUSIM Techniques and its Components in a Cloud Computing Environment”, International Journal of Computer Trends and Technology (IJCTT), Vol. 4, No. 8, Pp. 2435–
2440.
[15] Dr. Pawan Kumar & Gaganjot Kaur , “Study of Comparison of Vari- ous Cloud Computing Simulators”, IITT College of Engineering & Tech- nology, 2nd National Conference in Intelligent Computing & Communication, GCET Greater Noida, India
[16] The Network Simulator Ns2,
http://www.isi.edu/nsnam/ns/.
[17] http://www.cloudbus.org/greencloud/
[18]Banks, J., Carson, S. II, and Nelson, B.L., 199Discrete-Event System Simulation, 2nd ed.,Prentice Hall, Upper Saddle River, N.J.
[19]Chang, Chris, 2004. Simulation modeling handbook: apracti- cal approach, CRC Press LLC, ISBN 0-8493-1241-8
Co-Author:Mr. Sandeep .Gonnade is working as an Asst.Professor in the Department of Computer science Engineer- ing, School of Engineering and IT,MATS UNIVERSITY ,RAIPUR (C.G.). He has published 12 papers in different journals. Email:sandeep_gonnade@yahoo.co.in,
IJSER © 2014 http://www.ijser.org