Study of building a framework for modeling and simulation of cloud computing (infrastructure and services)
Prepared by the researcher : Dr. Salma Othman Muhammad Qismallah, Assistant Professor of Computer Science, Faculty of Computer Science and Information Technology, Sudan Open University, Sudan
Democratic Arabic Center
Journal of Afro-Asian Studies : Nineteenth Issue – November 2023
A Periodical International Journal published by the “Democratic Arab Center” Germany – Berlin
:To download the pdf version of the research papers, please visit the following link
Abstract
The study aims to build a framework for modeling cloud architectures for electronic clouds, define an approach that works efficiently and effectively to simulate infrastructures, and compare the results with a real cloud environment to improve services, because experiments in a real cloud environment are very difficult and expensive, which produces unexpected reports in the implementation phase. It is difficult to solve in order to perform operations again and determine what is required. The importance of the study lies in the possibility of expansion, development, building and testing of technologies to improve services, for the total adoption of this technology by institutions. The study used the applied descriptive analytical approach, by creating and preparing data centers and the hosts, hosts and virtual machines they contain. And prepare cloud tasks and connect them to virtual machines, and determine the start time of the cloud simulation and the end time of the simulation, all of these operations using common, slips, and java, one of the most important results of this study is the presence of different information to measure capabilities with tasks and cases and the time and end of the simulation, and changing the number of hosts and displaying these experiments in the cloud before uploading them to the real world, which reduces energy consumption and finding solutions that leads to increased data efficiency. One of the most important recommendations of the study is to focus on developing work on Data centers and linking them with artificial intelligence and deep learning techniques, and benefiting from data mining and forecasting tools.
introduction:
The use of modern information technologies has become a necessity, for reasons including the continued high storage costs, the difficulty of retrieving data, the number of backup copies, and effective control in controlling the costs of hardware, software, etc. Therefore, cloud computing focuses on providing a reliable, secure, fault-tolerant, sustainable, and scalable infrastructure to host Internet-based application services and these applications have a wide configuration in different deployment requirements.
Capture One: Methodology framework:
Study problem:
Experiments in the real environment or real reality are very difficult and expensive and therefore require extracting unexpected reports during the implementation stage. Therefore, they are difficult to solve and require an expensive cost to perform the operations again and correct what is required.
There is a very high power consumption, which affects dynamic provisioning, multi-tenancy, server utilization and data center efficiency.
Importance of the study:
– Contributing to the attempt to identify cloud computing, the near and long-term future of this technology, and the problems facing its applications, especially in developing countries.
-The reliance of a large number of global institutions on cloud computing services.
-Possibility of expansion and development in response to changing application requirements.
-Build and test new technologies to improve services.
-Ease of evaluating the researchers’ usage method before actual use and comparing the results in different conditions compared to the real cloud environment.
-Adjust input parameters.
Study objectives:
1-Building a new framework for modeling and simulation of infrastructure and cloud services.
2- Leverage appropriate current simulation tool technologies and identify an approach that works effectively to simulate and model cloud infrastructure and services and reduce overall server energy consumption and data center efficiency.
3- Conduct preliminary experiments using technical simulation tools and show the final image in a scenario before using it in the real world.
4-Compare the results with the real cloud environment and test new techniques to improve services.
Study hypotheses and suggestions:
1-Instead of storing software applications and data locally on a personal computer (IA as, SaaS, Peas, for example).
2-In this study, we propose a new generalized and scalable simulation framework that enables the modeling, simulation and experimentation of emerging computing infrastructures and management services.
Capture Tow: Theoretical framework:
The electronic cloud is the main trend currently, and it is a real trend in various fields in the form of software and information technology, and combining this data in one platform that is saved on the Internet and provides needs upon request, and provides users’ resources at any time in a dynamic way, which leads to lower costs and focus on basic tasks. Providing services in all fields.
Cloud computing provides IT infrastructure and applications as services to end users under a usage-based payment model, and virtual services can be leveraged based on time-varying service requirements.1
There are difficulties in provisioning, configuring and deploying the performance of cloud provisioning policies and the load of application operations, and managing repeatable resources under different system and user configurations and requirements that are difficult to achieve. To overcome these challenges, we propose using the Clouds tool, which is an extensible simulation toolkit that allows modeling the system and behavior of the system components that It includes data centers, virtual devices, and resource provisioning policies. This tool displays custom interfaces for implementing policies and service provisioning techniques to reduce virtual machines within the cloud overlapping networks. It works to implement general application techniques that can be easily expanded with limited effort, and the ability to easily reproduce results and ensure accuracy and visualization. In virtual reality before uploading it to real reality.
Clouds enables greater focus on innovation and business value creation.
Cloud SIM:
2It is a new scalable simulation framework that allows modeling, simulation and experimentation of emerging cloud computing infrastructures and application services.
It is a stand-alone platform developed by a laboratory at the University of Melbourne to work on simulating and experimenting3 on seamless infrastructure design. It was built on top of the grid computing framework that was developed by the same laboratory. GRID
This tool offers great advantages to customers and service providers, as it allows them to choose their services in a controlled environment4 at no cost, verify performance before deploying on the real cloud, and improve the cost of accessing resources while improving profits. Without these tools, the evaluation of customers and providers would be inaccurate and subject to trial and error. Which leads to ineffective service performance and reduced revenue generation, the CloudSIM tool helps researchers and developers in the industry to test the performance of an advanced application service in a suitable and easy-to-set-up environment.
The advantages of using CloudSIM for initial performance testing are: 1-2
Time efficiency (it takes less time and effort to bring work to the cloud and applications)
Flexibility.5 (the ability for developers to easily model and test the performance of their applications and services in heterogeneous environments.
[Microsoft Azure, Amazon EC2]6
1-Support unified cloud environment simulation.
2-Support simulating network communications between components of the simulated system.
3- A cloud platform that includes all services.
4-Independent and shared virtual services hosted on the data center node.
5- Support and simulate cloud on a large scale (computing environments, including data centers).
The cloud platform provides services on a subscription basis in a pay-as-you-go model to cloud customers, and therefore cloud infrastructure modeling and simulation tools must be available7 to support economic entities such as cloud brokers and cloud exchanges to enable real-time trading of services between customers and providers, among these The available simulations are the final visualization that provides support for managing economically driven resources and simulating application scheduling. We used this tool as a main element in this study in order to achieve the cloud goals of providing virtual resources for incoming user requests, application scheduling, resource discovery, and negotiations between the cloud and cloud federation in order to support and accelerate applications. Cloud services are a necessity and the appropriate software tool is designed to assist and develop researchers and developers.
The benefit of CloudSIM lies in the dynamic provisioning of application services.8 In a hybrid cloud environment, researchers and developers have found that this tool fulfills their demands for cloud resource provisioning and energy-efficient management of data center resources.
CloudSIM allows the development of best practices and processes in all critical aspects related to cloud computing and is able to test failure and recovery mechanisms.
CloudSIM provides9 virtualization engine to help create and manage multiple, independent, and coordinated virtual services on data center nodes and flexibly switch between time sharing and space sharing when allocating processing cores to virtual services. [
10Benefits of Cloudsim Tools:
Cloud market modeling.
Network behavior modeling.
Cloud federation modeling.
Modeling the cloud computing environment.
Modeling virtual machine allocation.
Dynamic workload applications are supported.
Supports GREEN IT policies.
Speed up application execution time.
Cloud SIM features the following:
1- Flexibility (adjust configuration).
2- Generate cloud parameters.
3- Save and easy access for all users.
4- Its lack of dependence on the system’s website.
5- Design and develop applications at low costs.
6- Evaluating the effectiveness of strategies in different scenarios
7- There are basic categories that build on Clouds
8- Creates tasks in the cloudlet environment.
11The data center in which virtual network resources are provided for processing and querying virtual machine information. Data Center
It hides the management of virtual machines such as creating, submitting tasks, and destroying virtual machines. Data Center Broker
Extends the parameter allocation strategy from machine to virtual machines, and one host can correspond to multiple virtual machines. Broker.
A virtual machine running on the host to share resources with other virtual machines. VirtualMachine.12
It means the scheduling strategy for virtual machines and works to manage execution tasks and implement task interfaces.VM Scheduler
Provides a description of the virtual machine.VM Characteristics
A virtual machine monitoring policy that describes the strategy for sharing resources for multiple virtual machines on the same host.VM M Allocation Policy.
Implement and assign data center hosts to VM Provisional virtual machines
Basic model of cloud simulator:
The electronic cloud consists of various components such as server, data center, virtual machines, users, adapters, and varying capabilities, and the simulation tool uses a set of operations such as scheduling, virtual machine migration, data and resource management, and security.
The simulator works in three stages:
First: The input is presented as a task or resource, and then the required operations are performed.
Second: Obtaining the result of the technology analysis.
Each simulator contains four components:
1- Application layer: allows users to send requests and responses.
2- The cloud virtualization layer that virtualizes cloud resources.
3- The cloud resource layer that includes hardware components such as memory, processor, and storage.
4 – Cloud simulation layer that provides libraries for simulation management.
Cloud Infrastructure
Figure No. (1) shows the components of the cloud
Capture Three: Methodology:
Cloudsim is one of the cloud simulation tools that has the following specifications:
The experiment was conducted on various equal and different parameter parameters (Time, Maps).
- The cloud was simulated in Sudan at the Virtual University and its branches, and it has a network within cities in 3 cities and a point of presence in 30 centers.
- Coding was done to model the data center with two servers. The number of VM is equal to the number of cities i.e. 3. However, the data center has enough RAM, bandwidth and storage specifications to accommodate data from 30 cores, the number of servers, RAM, bandwidth and storage specifications can be increased as per the demand.
- Tasks are fetched as small cloudlets from thin clients, so a minimum configuration of cloudlet length and file size is assumed.
- Simulation was performed to accommodate cloudlets applications from a maximum of 110 cities/centers, and VM image size, VM RAM, VM MIPS, VM bandwidth and cloud length parameters were changed accordingly.
- The above-mentioned differences in two virtual machine provisioning policies: shared-time and shared-space are studied.
- The effect of the above parameter variations on the processing cost and cloud completion time was measured.
Main activities of the cloud simulation tool:
| Activity | Description of the activity
|
| The definition | describes the process in detail to understand the cloud environment.
|
| Accurate | identification highlights the factor that affects performance and cost.
|
| Operations | supports different configurations to represent real-world conditions.
|
| The analysis | is able to evaluate the diverse requirements of cloud customers.
|
| Scale | allows researchers to run algorithms at different levels to validate the technique.
|
| Confirm | the use or modification of technical reports in a scenario before using them in the real world.
|
A table showing the main activities of the simulation tool
The applied descriptive analytical approach from the CIOUDSIM simulation tool was used, which is to enter data into the data center simulator, and then extract reports that show the allocation of shared space and the allocation of shared time to the processing core to the virtual services and resources available in the data center.
Mechanism of action:
We have a Data Center database that contains hosting and a number of virtual machines. The cloud simulation tool took several steps to operate the cloud and display reports before uploading them to the real world, which are:
The first step:
Initialize the Cloud Tool before creating any object
// First step: Initialize the Clouds package. It should be called
// before creating any entities.
intnum_user = 1; // number of grid users
Calendar = Calendar.getInstance();
booleantrace_flag = false; // mean trace events
// Initialize the CloudSim library
CloudSim.init(num_user, calendar, trace flag);
The second step:
Creating a data center, which requires initializing all objects that build on the data center
// Second step: Create Datacenters
//Datacenters are the resource providers in CloudSim. We need at list one of them to run a CloudSim simulation
@Suppress Warnings(“unused”)
Datacenter datacenter0 = create Datacenter(“Datacenter_0”);
@Suppress Warnings(“unused”)
Datacenter datacenter1 = create Datacenter(“Datacenter_1”);
The third step:
Create objects that build on the data center (host)
//Third step: Create Broker
Datacenter Broker = create Broker ();
intbrokerId = broker. get ID ();
The fourth step:
Creating and configuring virtual machines
//VM Parameters
long size = 10000; //image size (MB)
in ram = 512; //vim memory (MB)
in maps =120000;
long bow = 1000;
intpesNumber = 1; //number of Cups
String vim = “Xin”; //VMM name
//Fourth step: Create VMs and Cloudlets and send them to broker
vmlist = create(brokerId,20); //creating 20 vims
cloudlet List = create Cloudlet (broker ID,12000); // creating 12000 cloudlets
broker.submitVmList(vmlist);
broker. submitCloudletList (cloudlet List);
Step 5:
Create cloud tasks
// Fifth step: Starts the simulation
CloudSim.startSimulation();
// Final step: Print results when simulation is over
List<Cloudlet>new List = broker. getCloudletReceivedList();
CloudSim.stopSimulation();
printCloudletList (new List);
Log.printLine(“CloudSimExample6 finished!”);
}
catch (Exception e)
{
- printStackTrace();
Log.printLine(“The simulation has been terminated due to an unexpected error”);
private static Datacenter create Datacenter (String name) {
// Here are the steps needed to create a Power Datacenter:
// 1. We need to create a list to store one or more
// Machines
List<Host>host List = new Array List<Host>();
// 2. A Machine contains one or more PEs or CPUs/Cores. Therefore, should
// create a list to store these PEs before creating
// a Machine.
List<Pe> peList1 = new Array List<Pe>();
intuits = 120000;
// 3. Create PEs and add these into the list.
//for a quad-core machine, a list of 4 PEs is required:
peList1.add (new Pe (0, new PeProvisionerSimple(mips))); // need to store Pe id and MIPS Rating
peList1.add (new Pe (1, new PeProvisionerSimple(mips)));
peList1.add (new Pe (2, new PeProvisionerSimple(mips)));
peList1.add (new Pe (3, new PeProvisionerSimple(mips)));
//Another list, for a dual-core machine
List<Pe> peList2 = new Array List<Pe> ();
peList2.add (new Pe (0, new PeProvisionerSimple(mips)));
peList2.add (new Pe (1, new PeProvisionerSimple(mips)));
//4. Create Hosts with its id and list of PEs and add them to the list of machines
inthostId=0;
in ram = 512; //host memory (MB)
long storage = 100000; //host storage
intbw = 1000;
hostList.add (
new Host (
hosted,
new RamProvisionerSimple(ram),
new BwProvisionerSimple(bw),
storage,
peList1,
new VmSchedulerTimeShared(peList1)
); // This is our first machine
hosted++;
hostList.add (
new Host (
hosted,
new RamProvisionerSimple(ram),
new BwProvisionerSimple(bw),
storage,
peList2,
new VmSchedulerTimeShared(peList2)
)
); // Second machine
//To create a host with a space-shared allocation policy for PEs to VMs:
//hostList.add(
Create a number of virtual machines and link cloud tasks with those machines
private static List<Vm>vmlist;
private static List<Vim>create (intuserId, intvms) {
//Creates a container to store VMs. This list is passed to the broker later
Linked List<Vm> list = new LinkedList<Vm> ();
//create VMs
Vm[] vm = new Vm[vms];
for(inti=0;i<vms;i++){
vm[i] = new Vm(i, userId, mips, pesNumber, ram, bw, size, vmm, new CloudletSchedulerTimeShared ());
//for creating a VM with a space shared scheduling policy for cloudlets:
//vm[i] = Vm(i, userId, mips, pesNumber, ram, bw, size, priority, vmm, new CloudletSchedulerSpaceShared());
list.add(vm[i]);
return list;
private static List<Cloudlet>createCloudlet(intuserId, int cloudlets){
// Creates a container to store Cloudlets
LinkedList<Cloudlet> list = new LinkedList<Cloudlet>();
//cloudlet parameters
long length = 1000;
long fileSize = 500;
long outputSize = 500;
intpesNumber = 1;
UtilizationModelutilizationModel = new UtilizationModelFull();
Cloudlet[] cloudlet = new Cloudlet[cloudlets];
for(inti=0;i<cloudlets;i++){
cloudlet[i] = new Cloudlet(i, length, pesNumber, fileSize, outputSize, utilization Model, utilization Model, utilization Model);
// setting the owner of these Cloudlets
cloudlet[i].setUserId(userId);
list.add(cloudlet[i]);
return list;
Results:
Code outputs for the first step of the third step:
Starting Cloud Sim version 3.0
Datacenter_0 is starting…
Datacenter_1 is starting…
Broker is starting…
Entities started.
Code output for step four:
: Broker: Cloud Resource List received with 2 resource(s)
0.0: Broker: Trying to Create VM #0 in Datacenter_0
0.0: Broker: Trying to Create VM #1 in Datacenter_0
0.0: Broker: Trying to Create VM #2 in Datacenter_0
0.0: Broker: Trying to Create VM #3 in Datacenter_0
0.0: Broker: Trying to Create VM #4 in Datacenter_0
0.0: Broker: Trying to Create VM #5 in Datacenter_0
0.0: Broker: Trying to Create VM #6 in Datacenter_0
0.1: Broker: VM #0 has been created in Datacenter #2, Host #0
0.1: Broker: VM #1 has been created in Datacenter #2, Host #0
0.1: Broker: VM #2 has been created in Datacenter #2, Host #0
0.1: Broker: VM #3 has been created in Datacenter #2, Host #1
0.1: Broker: VM #4 has been created in Datacenter #2, Host #0
0.1: Broker: VM #5 has been created in Datacenter #2, Host #1
Outputs of the fifth and sixth steps:
Broker is shutting down…
Simulation: No more future events
CloudInformationService: Notify all CloudSim entities for shutting down.
Datacenter_0 is shutting down…
Datacenter_1 is shutting down…
Broker is shutting down…
Simulation completed.
Simulation completed.
tarting CloudSimExample6…
Initialising…
Simulation completed.
CloudSim finished
========== Out Put ==========
Cloudlet ID STATUS Data center ID VM ID Time Start Time Finish Time
4 SUCCESS 2 4 0.31 0.2 0.51
16 SUCCESS 2 4 0.31 0.2 0.51
28 SUCCESS 2 4 0.31 0.2 0.51
5 SUCCESS 2 5 0.31 0.2 0.51
17 SUCCESS 2 5 0.31 0.2 0.51
29 SUCCESS 2 5 0.31 0.2 0.51
6 SUCCESS 3 6 0.31 0.2 0.51
18 SUCCESS 3 6 0.31 0.2 0.51
30 SUCCESS 3 6 0.31 0.2 0.51
7 SUCCESS 3 7 0.31 0.2 0.51
19 SUCCESS 3 7 0.31 0.2 0.51
31 SUCCESS 3 7 0.31 0.2 0.51
8 SUCCESS 3 8 0.31 0.2 0.51
20 SUCCESS 3 8 0.31 0.2 0.51
32 SUCCESS 3 8 0.31 0.2 0.51
10 SUCCESS 3 10 0.31 0.2 0.51
22 SUCCESS 3 10 0.31 0.2 0.51
34 SUCCESS 3 10 0.31 0.2 0.51
9 SUCCESS 3 9 0.31 0.2 0.51
21 SUCCESS 3 9 0.31 0.2 0.51
33 SUCCESS 3 9 0.31 0.2 0.51
11 SUCCESS 3 11 0.31 0.2 0.51
23 SUCCESS 3 11 0.31 0.2 0.51
35 SUCCESS 3 11 0.31 0.2 0.51
0 SUCCESS 2 0 0.42 0.2 0.62
12 SUCCESS 2 0 0.42 0.2 0.62
24 SUCCESS 2 0 0.42 0.2 0.62
36 SUCCESS 2 0 0.42 0.2 0.62
1 SUCCESS 2 1 0.42 0.2 0.62
13 SUCCESS 2 1 0.42 0.2 0.62
25 SUCCESS 2 1 0.42 0.2 0.62
37 SUCCESS 2 1 0.42 0.2 0.62
2 SUCCESS 2 2 0.42 0.2 0.62
14 SUCCESS 2 2 0.42 0.2 0.62
26 SUCCESS 2 2 0.42 0.2 0.62
38 SUCCESS 2 2 0.42 0.2 0.62
3 SUCCESS 2 3 0.42 0.2 0.62
15 SUCCESS 2 3 0.42 0.2 0.62
27 SUCCESS 2 3 0.42 0.2 0.62
39 SUCCESS 2 3 0.42 0.2 0.62
Capture Four: Conclusions and results:
1-A model framework for simulating cloud architecture has been proposed.
2- There are differences in comparing information to measure capabilities depending on tasks, situations, start time, end time, and changing the number of hosts at the end of the given time.
3- Displaying initial experiments before uploading them to the actual reality, which leads to reduced energy consumption and efficiency of data centers by developing solutions before uploading them to the platform.
Recommendations:
1-Focusing on the efficiency of data centers and developing simulation tools to solve energy consumption problems
2-Connecting the CloudSim tool with artificial intelligence techniques and deep learning tools
3- Utilizing the CloudSim tool in data mining and making predictions and forecasts of events.
List of Reference:
- Armbrust, A. Fox, R. Griffith, A. Joseph, R. Katz, A. Konwinski, G. Lee, D. Patterson, A.Rabkin, I. Stoica,M. Zaharia. Above the Clouds: A Berkeley View of Cloud
computing. Technical Report No. UCB/EECS-2009-28, University of California at Berkley, USA, Feb. 10, 2009
. 2-CloudAnalyst: A Cloud Sim-based Tool for Modeling and Analysis of Large Scale Cloud Computing Environments, ME Project Report Bhathiy aWickremasinghe
(mkbw@pgrad.unimelb.edu.au)Student No: 31828222/6/2009, Project Supervisor :Assoc. Prof. Rajkumar Buyya.
3-CloudSim: a toolkit for modeling and simulation of cloud computing environments and evaluation of resource provisioning algorithms ,Rodrigo N. Calheiros, Rajiv Ranjan, Anton Beloglazov, César A. F. De Rose, Raj Kumar Buyya,First published: 24 August 2010
4-Experimental comparison of simulation tools for efficient cloud and mobile cloud computing application Khadijah Bahwaireth1, Lo’ai Tawalbeh1, , Elhadj Benkhelifa, Yasser Jararweh2*and Mohammad A. Tawalbeh, Bahwairethet al. EURASIP Journal on Information Security (2016) 2016:15 DOI 10.1186/s13635-016-0039-y.
5-Armbrust M, Fox A, Griffith R, Joseph A, Katz R, Kaminski Alee G, Patterson D, Rabin A, Stoical I, Zaharia M. A view of cloud computing. Communications of the ACM 2010; 53(4):50–58.
6-Buyya R, Yeo CS, Venugopal S, Broberg J, Brandic I. Cloud computing and emerging IT platforms: Vision, hype, and reality for delivering computing as the 5th utility. Future Generation Computer Systems 2009; 25(6):599–616.
7-A Simplified Model for Simulating the Execution of a Workflow in Cloud- Euro-Par 2017: Parallel Processing, Springer, Cham.
8-Simulation Modeling Practice and Theory Volume 93, May 2019, Pages 133-147.
9-F. Howell and R. McCabe. Sim Java: A discrete event simulation library for java. In Proceedings of the first International Conference on Web-Based Modeling and Simulation, 1998.
10-J. Byrne et al., “A Review of Cloud Computing Simulation Platforms & Related Environments,”
Proc. 7th Int. Conf. Cloud Compute. Serv. Sci., no. Closer, pp. 651–663, 2017, doe: 10.5220/0006373006790691.
11-V. Thamar, “A Comparative Study of Cloud Simulation Tools,” Int. J. Compute. Sci. Eng., vol. 9, no.
06, pp. 385–392, 2017
12- S. VA hora and R. Patel, “Cloud Sim-A Survey on VM Management Techniques,” Int. J. Adv. Res.
Compute. Common. Eng., vol. 4, no. 1, pp. 128–133, 2015, doi:
10.17148/ijarcce.2015.4126.
