Section A

Comparison of Cloud platforms

AWS offers users a pay-as-you-go approach for cloud service that is over 160 whereby one pays individual services needed upon using them with no long term contract or any licensing. Its offers a storage, analytics, database, application, global compute and deployment services that aid organizations lower IT costs, move faster and scale up applications [1[. The services are priced uniquely with no complex dependencies or any license requirement. AWS provides a free usage service to new customers to help them get started in cloud platform. The free service enables the user run anything s/he wants like testing existing applications, launching new applications or any gaining a hands on experience on use of AWS.

Microsoft Azure cloud service providers use a built in billing and a subscription model on pay per use basis such that if you use two hours CPU time you only pay for the two hours consumed. It offers free security when migrating windows server to Azure virtual machine [2]. Saving occurs  when compute capacity is paid on one or three year term with a reservation price that improves budget forecasting providing flexible exchange when business need change or upon cancelling. The hybrid gives more value from windows server licences with a 40 % save on the virtual machines.

  Google cloud platform offers zero up front cost and pay as you go price with no termination fee whereby upon turning off a service, you also stop paying for it [1]. It offers an increased 30 percent discount work that run for significant monthly billing portion on computing engine or cloud SQL. In addition it provides a 79 percent off on interrupted workloads such as data processing and data mining. Picking a given CPU and memory configuration saves 48 percent in comparison to fixed machine types.

Open stack platform does not offer free trial but keeps cost under control by making a payment for only what one uses up to the minute. It uses a true consumption based pricing and billing strategy. The computing service adds up for cloud block storage. The price per hour increases as the network, CPU weight and RAM services increase

Simulation tools available in cloud computing

Cloud computing is a modern technology that allows the computer users to share data over the internet by purchasing the service from third party companies. The service is made possible through advanced use of virtualization technologies and distributed computing. The Cloud computing companies offer the cloud services using various essential models including the Unified Communications as a Service (UCaaS), Software as a service (SaaS), Platform as a service (PaaS), and Infrastructure as a service (IaaS) [3]. The approaches make the cloud computing agile, while still achieving cost efficiency, reduced error rate, increased reliability and location flexibility. Some cloud computing services include the real-time data processing, content delivery, web hosting, social networking, among other services. Allocation policies and in real-time computing and quantifying the cost of the performance of various schedules is costly and challenging, thus require cloud simulation tools to assess the performance of the cloud systems with ease. The cloud simulations include NetworkClouSim, GroudSim, Open Cirrus, Open Cloud Testbed (OCT), GreenCloud, SPECI and CloudSim [4].

CloudSim is a library developed in Australia for simulation of Cloud computing services. The tool offers fundamental classes for describing data centres, computational resources, users, applications, virtual machines and policies to manage different parts of the system [5]. The components can be combined for the users to perform mapping, schedule algorithms, evaluate new policies, among other functions. CloudSim allows complex simulations by replacing or extending classes or creating new codes foe the desired situations.

CloudSim is not a ready to use solution where the user can just set parameters and derive the results in the project. However, it is a library where the user needs to write Java code to create the desired results, thus assess security and performance of Cloud applications [6]. The elements in the CloudSim library communicate with each other through data packets allowing the user to model data centres, allocate virtual machines, assess network behaviors and power consumption. There are other simulation tools which imitates the CloudSim technology including CloudAnalyst, CloudMIG Xpress, Cloud Auction, CloudReports, RealCloudSim, SimpleWorkflow, WorkflowSIm and CloudSimEx.

Simulation Program for Elastic Cloud Infrastructures (SPECI) is a simulation tool that enables the user to explore the aspects of scaling and performance of prospect data centres.  SPECI simulates the behavior and performance of data centres. GreenCloud is a packet-level cloud simulation tool which assesses the energy consumption of the data centres and IT components in cloud computing [6]. The tool is energy aware thus is useful in developing novel solutions like optimization of communication protocols, workload scheduling, resource allocation, monitoring and optimization of network infrastructures.

Open Cloud Testbed (OCT) cloud computing simulation tool benchmarks several cloud computing systems to assess their interoperability. The tool s used as a small-scale testbed for developing cloud-based infrastructure and software. OCT architecture allows high-performance services, protocols and infrastructure which allow high speed of connectivity and transfer of protocols. The OCT tool simulates several cloud systems to make it easier to analyze the interoperability and achieve benchmarking [7]. Additionally, the different cloud systems integration helps in developing network libraries and monitor benchmarking suites and systems to help in developing and experimenting for different cloud computing stacks.

The Open Cirrus cloud simulation tool aims at developing open-source APIs and stacks for the cloud and offering a group of experimental datasets. Additionally, the Open Cirrus encourages application-level research and new cloud computing applications, thus fostering systems-level research in cloud computing. The Open Cirrus has both virtual and physical machines and international services including job submission, storage, monitoring and sign-on, thus offering a platform for real-world services and applications. NetworkCloudSim is an extension of CloudSim but with enhanced network scalability and universal application model. NetworkCloudSim allows accurate assessment of resource provision policies and scheduling for Cloud performance optimization.

GroudSim is an event-based cloud simulation tool that offer a comprehensive combination of features for complex simulation work [6]. The simulation can be parameterized and extended through probability distribution packages for errors that may happen in complex cloud environments. GroundSim operates using Iaas but may be extended to support other models including Taas, DaaS and PaaS. The cloud simulation tools are chosen based on the user requirements.

Graph Theory Usefulness in cloud computing

Graph based representation provides a simplified view of a complex network while graph technique simplifies solutions for inherent issues in networks. Graphs keep on increasing in size with a seemingly unending number of facts and the details persons would be interested to know about each other. An example is a geographical location [7]. A simplified standard map – a graph – provides the short route between any two cities. However, more detailed analysis can be done to provide more details including weather conditions, road-works, expected traffic jams, speed limits, among other aspects. The theory can help learn the most fuel-efficient route or the most scenic route or one which has the most convenient rest sections [8]. These type of details information can be retrieved from the map when the user has the right inputs and tools.

The graph theory helps in proving graph-based data presentations to simplify the complex data from different sources [9]. Thus, graph theory helps in achieving greater ease of use of many issues in networks through simulation of multi data sources and simplifying through graphical presentation.

Section B

Setting up network cluster on virtualBox

Properties of the Host machine

Fig 1: Host machine properties

Downloading virtual Box and installing the virtual Box on the Host machine

Fig 2: Installed Virtual-Box

Creating a new machine

Created virtual machines on virtual Box

Master virtual machine: Vulnerability scanner

Other virtual machine act as cluster network

Generating random Traffic

Generating random traffic is the same as network stressing. This is a procedure of finding out whether a computer or any other application can withstand high loads and at the same time remain operational

1^st procedure

Sending large and unstoppable packets to a target virtual machine (192.168.100.3). The command ping in this case is –s 8500

Using network stressing tool nping

Installing nping

Using nping

Data manipulation 

Data manipulation in cloud is used to update, delete, and insert. In cloud environment data manipulation is the process of analyzing, visualizing and aggregating data. Usually, a user incorporates visualizations along with published reporting and visual analysis. To showcase this, this report has utilized the use of mkdir command as shown below.

Displaying home directory path

Creating new directories

Creating directories

2^nd directory creation

Navigating directory contents

Copying sample file to home and displaying the contents

References

[1] “Pricing.” Amazon Web Services, Inc, aws.amazon.com/pricing/.

[2] “Pricing.” Rackspace, www.rackspace.com/openstack/public/pricing.

[3] “Microsoft Azure Pricing and Licensing: 6 Things You Should Know.” Sherweb, 28 June 2018, www.sherweb.com/blog/cloud-server/understanding-microsoft-azure-pricing/.

 [4] Fakhfakh, Fairouz, Hatem Hadj Kacem, and Ahmed Hadj Kacem. “Simulation tools for cloud computing: A survey and comparative study.” 2017 IEEE/ACIS 16th International Conference on Computer and Information Science (ICIS). IEEE, 2017.

[5] Alshammari, Dhahi, Jeremy Singer, and Timothy Storer. “Performance evaluation of cloud computing simulation tools.” 2018 IEEE 3rd International Conference on Cloud Computing and Big Data Analysis (ICCCBDA). IEEE, 2018.

[6] Bahwaireth, Khadijah, et al. “Experimental comparison of simulation tools for efficient cloud and mobile cloud computing applications.” EURASIP Journal on Information Security 2016.1 (2016): 15.

[7] Devitt, Simon J. “Performing quantum computing experiments in the cloud.” Physical Review A 94.3 (2016): 032329.

[8] Sujan, S., and R. Kanniga Devi. “An efficient task scheduling scheme in cloud computing using graph theory.” Proceedings of the International Conference on Soft Computing Systems. Springer, New Delhi, 2016.

[9] Zhang, D. G., Tang, Y. M., Cui, Y. Y., Gao, J. X., Liu, X. H., & Zhang, T. (2018). Novel reliable routing method for engineering of internet of vehicles based on graph theory. Engineering Computations.