a load balancing model based on cloud partitioning for the public cloud

1

A Load Balancing Model Based on Cloud Partitioning for the Public Cloud

INDIRA GANDHI NATIONAL OPEN UNIVERSITY

A Load Balancing Model Based on Cloud Partitioning

For the Public Cloud

2



INTRODUCTION:

Cloud computing is an attracting technology in the field of computer science. In

Gartners report, it says that the cloud will bring changes to the IT industry. The cloud is

changing our life by providing users with new types of services. Users get service from a

cloud without paying attention to the details. NIST gave a definition of cloud computing

as a model for enabling ubiquitous, convenient, on-demand network access to a shared

pool of configurable computing resources (e.g., networks, servers, storage, applications,

and services) that can be rapidly provisioned and released with minimal management

effort or service provider interaction. More and more people pay attention to cloud

computing. Cloud computing is efficient and scalable but maintaining the stability of

processing so many jobs in the cloud computing environment is a very complex problem

with load balancing receiving much attention for researchers.

Since the job arrival pattern is not predictable and the capacities of each node in

the cloud differ, for load balancing problem, workload control is crucial to improve

system performance and maintain stability. Load balancing schemes depending on

whether the system dynamics are important can be either static or dynamic. Static

schemes do not use the system information and are less complex while dynamic schemes

will bring additional costs for the system but can change as the system status changes. A

dynamic scheme is used here for its flexibility. The model has a main controller and

balancers to gather and analyze the information. Thus, the dynamic control has little

influence on the other working nodes. The system status then provides a basis for

choosing the right load balancing strategy.

The load balancing model given in this article is aimed at the public cloud which

has numerous nodes with distributed computing resources in many different geographic

locations. Thus, this model divides the public cloud into several cloud partitions. When

the environment is very large and complex, these divisions simplify the load balancing.

The cloud has a main controller that chooses the suitable partitions for arriving jobs while

the balancer for each cloud partition chooses the best load balancing strategy

3



OBJECTIVES:

Resource allocation and job scheduling are the core functions of cloud computing.

These functions are based on adequate information of available resources. Timely

acquiring resource status information is of great importance in ensuring overall

performance of cloud computing. This work aims at building a distributed system for

cloud resource monitoring and prediction. In this paper, we present the design and

evaluation of system architecture for cloud resource monitoring and prediction. We

discuss the key issues for system implementation, including machine learning-based

methodologies for modelling and optimization of resource prediction models. Evaluations

are performed on a prototype system. Our experimental results indicate that the efficiency

and accuracy of our system meet the demand of online system for cloud resource

monitoring and prediction

PROJECT CATEGORY:

NETWORKING-CLOUD

TOOLS/PLATFORM:

The major tools used for the development are as follows:

JAVA EE

Java Platform, Enterprise Edition (Java EE) is the standard in community-driven

enterprise software. Each release integrates new features that align with industry needs,

improves application portability, and increases developer productivity. Java EE offers a

rich enterprise software platform. The platform provides an API and runtime environment

for developing and running enterprise software, including network and web services, and

other large-scale, multi- tiered, scalable, reliable, and secure network applications. Java

EE extends the Java Platform, Standard Edition (Java SE), providing an API for object-

4



relational mapping, distributed and multi-tier architectures, and web services. The

platform incorporates a design based largely on modular components running on an

application server. Software for Java EE is primarily developed in the Java programming

language.

Java EE includes several API specifications, such as RMI, e-mail, web services,

XML, etc., and defines how to coordinate them. Java EE also features some

specifications unique to Java EE for components. These include Enterprise JavaBeans,

connectors, servlets, Java Server Pages (JSP) and several web service technologies. This

allows developers to create enterprise applications that are portable and scalable, and that

integrate with legacy technologies. A Java EE application server can handle transactions,

security, scalability, concurrency and management of the components it is deploying, in

order to enable developers to concentrate more on the business logic of the components

rather than on infrastructure and integration tasks.

Java Platform, Enterprise Edition (Java EE) is the standard in community-driven

enterprise software. Today Java EE offers a rich enterprise software p latform, and

with over 20 compliant Java EE 6 implementations to choose from, low risk and plenty of

options as the industry begins the rapid adoption of Java EE 7, work has begun on Java

EE 8. With a survey that received over 4,500 responses, the community has prioritized

the desired features for Java EE 8. In fact, the following JSRs have already been

submitted.

JSR 366 - Java EE 8

JSR 367 - The Java API for JSON Binding

JSR 368 - Java Message Service 2.1

JSR 369 - Java Servlet 4.0

JSR 370 - Java API for RESTful Web Services 2.1

JSR 371 - Model-View-Controller 1.0

JSR 372 - Java Server Faces 2.3

JSR 373 - Java EE Management API 1.0

JSR 374 - Java API for JSON Processing 1.1

5



Some of the key features and services of J2EE:

At the client tier, J2EE supports pure HTML, as well as Java applets or

applications. It relies on Java Server Pages and servlet code to create HTML or

other formatted data for the client.

Enterprise JavaBeans (EJBs) provide another layer where the platform's logic is

stored. An EJB server provides functions such as threading, concurrency, security

and memory management. These services are transparent to the author.

Java Database Connectivity (JDBC), which is the Java equivalent to ODBC, is the

standard interface for Java databases.

The Java servlet API enhances consistency for developers without requiring a

graphical user interface.

MYSQL

MySQL is the worlds most popular open source database, enabling the cost-

effective delivery of reliable, high-performance and scalable Web-based and embedded

database applications. The MySQL software delivers a very fast, multi-threaded, multi-

user, and robust SQL (Structured Query Language) database server. MySQL Server is

intended for mission-critical, heavy-load production systems as well as for embedding

into mass-deployed software.

MySQL is a database management system

A database is a structured collection of data. It may be anything from a simple

shopping list to a picture gallery or the vast amounts of information in a corporate

network. To add, access, and process data stored in a computer database, you

need a database management system such as MySQL Server.

MySQL databases are relational

A relational database stores data in separate tables rather than putting all the data

in one big storeroom. The database structures are organized into physical files

optimized for speed. The logical model, with objects such as databases, tables,

views, rows, and columns, offers a flexible programming environment. You set

6



up rules governing the relationships between different data fields, such as one-to-

one, one-to-many, unique, required or optional, and pointers between different

tables. The database enforces these rules, so that with a well-designed database,

your application never sees inconsistent, duplicate, orphan, out-of-date, or

missing data.

MySQL software is Open Source

Open Source means that it is possible for anyone to use and modify the software.

Anybody can download the MySQL software from the Internet and use it without

paying anything.

The MySQL Database Server is very fast, reliable, scalable, and easy to use

MySQL Server can run comfortably on a desktop or laptop, alongside your other

applications, web servers, and so on, requiring little or no attentio n. Its

connectivity, speed, and security make MySQL Server highly suited for

accessing databases on the Internet.

MySQL Server works in client/server or embedded systems

The MySQL Database Software is a client/server system that consists of a multi-

threaded SQL server that supports different back ends, several different client

programs and libraries, administrative tools, and a wide range of application

programming interfaces (APIs).

APACHE WEB SERVER

Apache Tomcat (or simply Tomcat, formerly also Jakarta Tomcat) is an open

source web server and servlet container developed by the Apache Software Foundation.

Tomcat implements the Java Servlet and the Java Server Pages (JSP) specifications from

Oracle, and provides a "pure Java" HTTP web server environment for Java code to run in.

In the simplest config, Tomcat runs in a single operating system process. The process

runs a Java virtual machine. Every single HTTP request from a browser to Tomcat is

processed in the Tomcat process in a separate thread. Apache Tomcat includes tools for

7



configuration and management, but can also be configured by editing XML configuration

files. Tomcat can be used stand-alone, but it is often used behind traditional web

servers such as Apache httpd, with the traditional server serving static pages and Tomcat

serving dynamic servlet and JSP requests. No matter what we call Tomcat, a Java servlet

container or servlet and JSP engine, we mean Tomcat provides an environment in which

servlets can run and JSP can be processed. Similarly, we can absolutely say a CGI-

enabled Web server is a CGI program container or engine since the server can

accommodate CGI programs and communicate with them according to CGI specification.

Between Tomcat and the servlets and JSP code residing on it, there is also a standard

regulating their interaction, servlet and JSP specification, which is in turn a part of Suns

J2EE (Java 2 Enterprise Edition). But what are servlets and JSP? Why do we need them?

Lets take a look at them in the following subsections before we cover them in much

more detail in the future.

Traditionally, before Java servlets, when we mention web applications, we mean a

collection of static HTML pages and a few CGI scripts to generate the dynamic content

portions of the web application, which were mostly written in C/C++ or Perl. Those CGI

scripts could be written in a platform- independent way, although they didnt need to be

(and for that reason often werent). Another approach to generating dynamic content is

web server modules. For instance, the Apache httpd web server allows dynamically

loadable modules to run on startup. These modules can answer on pre-configured HTTP

request patterns, sending dynamic content to the HTTP client/browser.. Now let us take a

look at the Java side. Java brought platform independence to the server, and Sun wanted

to leverage that capability as part of the solution toward a fast and platform independent

web application standard. The other part of this solution was Java servlets. The idea

behind servlets was to use Javas simple and powerful multithreading to answer requests

without starting new processes. You can now write a servlet-based web application, move

it from one servlet container to another or from one computer architecture to another, and

run it without any change (in fact, without even recompiling any of its code).

8



REQUIREMENT SPECIFICATON:

Hardware Requirement Specification

Server:

Processor : Intel Pentium

RAM : 2GB

Hard Disk : 80GB

CPU Speed : 2.60GHz

Monitor : VGA Color

Client:

Processor : Pentium

RAM : 1GB

Hard Disk : 40GB

CPU Speed : 1.60GHz

Monitor : VGA Color

Software Requirement Specification

Web Server : Apache Tomcat

Tools : J2EE

Front End : JSP

Language : Java

Back End : MySQL

Web Technology : HTML, XML, JavaScript

Operating System : Windows 7 /above

9



PROBLEM DEFINITION:

A locally distributed system has various computers interconnected by a local

communication network .In cloud computing, controlling access to information is

difficult. Also the user does not know where exactly data is stored .If in cloud computing

the data is stored as distributed manner. It is saved on remote location or virtual locations

randomly.

If it is going to upload the data randomly on cloud it leads to the imbalance in

cloud server storage. The status of some nodes of cloud may be overloaded while some

maybe in idle or normal state. We propose an efficient technique for balancing the load in

clouds. Load balancing is a process in which workload is distributed among nodes of the

system. It helps to improve resource utilization and response time. In case of extensive

environments the load balancing is carried out by creating partitions. The best load

balancing strategy helps to select the appropriate partition. The main controller performs

the load balancing operations.

EXISTING SYSTEM

A growing number of companies have to process huge amounts of data in a cost-

efficient manner. Classic representatives for these companies are operators of Internet

search engines. The vast amount of data they have to deal with every day has made

traditional database solutions prohibitively

Expensive .Instead, these companies have popularized an architectural paradigm

based on a large number of commodity servers. Problems like processing crawled

documents or regenerating a web index are split into several independent subtasks,

distributed among the available nodes, and computed in parallel.

10



PROPOSED SYSTEM

In recent years a variety of systems to facilitate MTC has been d eveloped.

Although these systems typically share common goals (e.g. to hide issues of parallelism

or fault tolerance), they aim at different fields of application. Map Reduce is designed to

run data analysis jobs on a large amount of data, which is expected to be stored across a

large set of share-nothing commodity servers.

Once a user has fit his program into the required map and reduce pattern, the

execution framework takes care of splitting the job into subtasks, distributing and

executing them. A single Map Reduce job always consists of a distinct map and reduce

program.

PLANNING AND SCHEDULING:

Gantt chart

Tasks Duration June July August September

weeks 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3

Project 4months

System study 2weeks

System analysis 4weeks

Design 3weeks

Coding and testing 4weeks

implementation 1week

11



PURPOSE AND SCOPE:

PURPOSE

There are several cloud computing categories with this work focused on a public

cloud. A public cloud is based on the standard cloud computing model, with service

provided by a service provider. A large public cloud will include many nodes and the

nodes in different geographical locations. Cloud partitioning is used to manage this large

cloud. A cloud partition is a subarea of the public cloud with divisions based on the

geographic locations. The architecture is shown in Fig.1The dotted line around the Blade

server/Top of rack (ToR) switch indicates an optional layer, depending on whether the

interconnect modules replace the ToR or add a tier .The load balancing strategy is based

on the cloud partitioning concept. After creating the cloud partitions, the load balancing

then starts: when a job arrives at Typical cloud partitioning. The system, with the main

controller deciding which cloud partition should receive the job. The partition load

balancer then decides how to assign the jobs to the nodes.

When the load status of a cloud partition is normal, this partitioning can be accomplished

locally. If the cloud partition load status is not normal, this job should be transferred to

another partition as server virtualization and client virtualization.

SCOPE AND BENEFIT

SERVER VIRTUALIZATION

.

CLIENT VIRTUALIZATION

PRACTICAL SOLUTIONS FOR OPTIMIZING TRAFIC FLOW

.IDENTIFY TRAFIC BOTTLENECK

VIRTUAL SWITCH ARCHITECTURES.

12



ANALYSIS:

USECASE DIAGRAM 1:

Browse

&encrypt

Upload

file

View and

load status

Data load balence

Allocate data

Block unauthorised

users

Update balancer

Authorise data

Receive data

Decrypted data

Store data

Cloud server

Destination

Cloud user

13



DATA FLOW DIAGRAM

The DFD (also known as bubble chart) is a simple graphical formalism that can

be used to represent a system in terms of the input data into the system, various process

carried on these data and the output data generated by the system.

The main reason why the DFD technique is so popular is because the fact that the

DFD is very simple formalism it is simple to understand use. A DFD is a very limited

number of primitive symbols to represent the functions performed by a system and the

data flow among the functions. A DFD model hierarchy represents various sub-

functions.

Data Flow :

A line with an arrow represents data flows. The arrow shows the direction of flow of data.

Process:

A Circle is used to despite a process. Processes are numbered and given a name.

Data Store :

A data store stores the data. Two parallel lines with square depict a data store. Processes may

store or retrieve data from a data store.

14



External Entity :

External Entities are represented by the rectangle, and are outside the system, such as venders

or customers wit that the system interacts. The designers have no control over them.

DATA FLOW DIAGRAMS

Context level DFD

First level DFD

Second level DFD

CONTEXT LEVEL DFD:

15



LEVEL 1 DFD

16



LEVEL 2 DFD

17



DFD:

18



ER DIAGRAM:

User_registration

Message

User_login

homework

homeworksubmit

Agent_login

Agent_reg

Grid Admin

Reg_mail

View & verify

Find_cloud

Work submits

or delete?

Work_submit

Work_as

sign

Agent_login

User_

login

Work assigned to

uname

password email

addres

s

uname

cpswrd password

name

email

uname

dept

status job

doc

job cost

cost feedback

secret key

paswd

unam

e

www

gname

dept pswrd

discipline

key

passwrd

uname

19



DATA DICTIONARY:

Table 1: Adminlogin

Descriptions: - This table keeps The Information about Admin

Table 2: File Upload

Descriptions:- This table keeps The Information about the uploaded file.

Table 3: Grid

Descriptions:- This table keeps The Information about grid.

Field Name Data type Constrain Size Description

Username Varchar Not null 13 Name of admin

Password Varchar Not null 13 Password of admin

Department Varchar Not null 20 Department of the admin

Discipline Varchar Not null 20 Discipline of the admin


username Varchar null 13 Name of job

Savefile Varchar null 60 location

depaertment Varchar null 20 Department name

year Int null Year of saving


gname Varchar null 20 Name of grid

www Varchar null 60 url address

Svm Varchar null 40 -

20



Table 4: Homework

Descriptions:- This table keeps The Information about kind of job.

Table 5: Homework_submit

Descriptions:- This table keeps The Information about job submission.


Username Varchar Null 30 Name of the job

feedback Varchar Null 80 Feedback from the user

history Varchar Null 80 Previous data

Cost Float Null Cost in Rupees

File Varchar Null 30 File name

doc Varchar Null 40 File location

Xyz Varchar Null 40 Resource1

win Varchar Null 40 Resource2

Winn Varchar Null 40 Resource 3


Username Varchar Null 70 Name of the job

Homeworkassign Varchar Null 70 Job Assignments

Department Varchar Null 20 Name of the department

Status Varchar Null 10 Delete/allocate

21



Table 6: Agent

Descriptions :- This table keeps The Information about Agent Registration.

Table 7: Agent_login

Descriptions:- This table keeps The Information about agent login details.


Username Varchar Null 20 Username of agent

Password Varchar Null 12 password

Confirmpassword Varchar Null 12 Password retype

Firstname Varchar Null 13 First name of agent

Lastname Varchar Null 13 Last name

Fathername Varchar Null 13 Name of father

Address Varchar Null 40 Address for

correspondence

Pincode Int Null Postal code

Phonenumber Varchar Null 11 Contact no

Email Varchar Null 20 Mail id

Discipline Varchar Null 20 Educational qualification

Department Varchar Null 20 Subject/dept

Secretkey Varchar Null 8 Key provided at the time of agent registration


Username Varchar null 20 username

Password Varchar null 13 Password

Secret key Varchar null 8 Key

22



Table 8: User_login

Descriptions:- This table keeps the Information about User login details.

TABLE 9: USER

Descriptions:- This table keeps The Information about User Registration.


Username Varchar Not null 20 username

Password Varchar Not null 13 password

Secret key Varchar Not null 8 key


Username Varchar Null 20 username

Password Varchar Null 13 password

Confirmpassword Varchar Null 13 Retyape password

Firstname Varchar Null 13 Name of user

Lastname Varchar Null 13 name

Fathername Varchar Null 20 Name of father

Address Varchar Null 40 address

Pincode Int Null Postal code

Phonenumber Varchar Null 11 Contact no

Email Varchar Null 20 Mail id

Discipline Varchar Null 20 education

Department Varchar Null 20 Department name

Secretkey Varchar Null 8 Key.

23



MODULES AND DESCRIPTION:

1. GRP-PSO Iteration

2. Resources Allocation

3. QOS calculation

4. Evaluation for Minimum Cost Value

1. GRP-PSO Iteration

After successful grid login, the user can choose what are the best resources are

available from the list of resources using GRP-PSO Technique. In Initial state all resource

particle represented as multiple yellow dot. All resources grid network connection is

design from weighted graph in data structure.

2. Resources Allocation

After the user booked the resources, the user has Shortest path algorithms use to

detect the best node based on QOS parameters like Virtual machine capability, user

friendly, latest version, high speed, improved accuracy, resource quality, Good

performance, previous history, power capability, backup capacity, grid network support,

customers current & previous feedback, policy of resource, debugging tools, cooperation

facility and cost. Based on the QOS (Quality of service with cost) of resource,GRP-PSO is

evaluating the all resources for your task. Suitable resources are find and the connected to

task submitting resource after n number of iteration process. Where, n = number of

available resources. The suitable resource is not available on the user specified date. If the

user likes, he can reschedule the resources.

3. QOS calculation

Minimum distance of x/y is use to find the best grid nodes from the lot of

available resources .Here x is a distance between the task submitting resource node to

other available resource node, y is a QOS rate. Where QOS rate is calculated based on

listed attributes. High quality attributes get high QOS rate. After the best grid node

24



founded transfer your tasks (job) from submitting node to best grid node through Globus

online grid network.

4. Evaluation for Minimum Cost Value

Minimum Cost value can be achieved through iteration using GRP-PSO

Algorithm. Finding the Minimum cost value to find the best node in grid Environment is found

through the Iterated Minimum cost value.

IMPLIMENTATION METHODOLOGY:

System implementation is the final phase i.e., putting the utility into action.

Implementation is the stage in the project where theoretical design turned in to working

system. The most crucial stage is achieving a new successful system and giving

confidence in new system that it will work efficiently and effectively. The system is

implemented only after thorough checking is done and if it is found working in according

to the specifications.

It involves careful planning, investigation of the current system and constrains on

implementation, design of methods to achieve. Two checking is done and if it is found

working according to the specification, major task of preparing the implementation are

educating, training the users.

IMPLEMENTATION PROCEDURES

The major implementation procedures are,

Test Plans

Training

Equipment Installation

Conversion

1. Test Plans:

The implementation of a computer-based system requires that test data be prepared and

that the system and its elements be tested in a planned, structured manner. The computer

25



program component is a major subsystem of the computer-based information system, and

particular attention should be given to the testing of this system element as it is

developed.

2. Training:

The purpose of the training is to ensures that all the personnel who are to be associated

with the computer-based business system possess the necessary knowledge skills,

Operating, Programming, and user personnel are trained using reference manuals as

training aids.

2.1 Programmer Training

2.2 Operator Training

2.3 User Training

3. Equipment Installation:

Equipment vendors can provide the specification for equipment installation. They usually

work with the projects equipment installation team in planning for adequate space,

power, and light and a suitable environment. After a suitable site has been completed, the

computer equipment can be installed. Although equipment normally is installed by the

manufacturer, the implementation team should advice and assist. Participation enables the

team to aid in the installation and, more importantly, to become familiar with the

equipment.

4. Conversion:

Conversion is the process of performing all of the operations that result directly in the

turnover of the new system to the user. Conversion has two parts:

1. The creation of a conversion plan at the start of the development phase and

the implementation of this plan throughout the development phase.

2. The creation of the system change over plan to the end of the development

phase and the implementation of the plan at the beginning of the operating

system.

26



ARCHITECTURE:

SECURITY MECHANISM:

The system security problem can be divided into four related issues: security,

integrity, privacy and confidentiality. They determine the file structure, data structure and

access procedures.

Cloud

User

Interne

t

Cloud

User

Interface

Admin

Cloud

Revelati

on

Evaluato

r

Cloud

Node-1

Cloud

Node-N

Revelatio

n Services

27



System security refers to the technical innovations and procedures applied to the

hardware and operating systems to protect against deliberate or accidental damage from a

defined threat. In contrast, data security is the protection of data from loss, disclosure,

modifications and destruction.

System integrity refers to the proper functioning of programs, appropriate

physical security and safety against external threats such as eavesdropping and

wiretapping. In comparison, data integrity makes sure that do not differ from original

from others and how the organization can be protected against unwelcome, unfair or

excessive dissemination of information about it.

The term confidentiality is a special status given to sensitive information in a data

base to minimize the possible invasion of privacy. It is an attribute of information that

characterizes its need for protection.

FUTURE ENHANCEMENT:

Since this work is just a conceptual framework, more work is needed to

implement the framework and resolve new problems. Some important points are:

Cloud division rules: Cloud division is not a simple problem. Thus, the framework

will need a detailed cloud division methodology. For example, nodes in a cluster may

be far from other nodes or there will be some clusters in the same geographic area

that are still far apart. The division rule should simply be based on the geographic

location (province or state).

How to set the refresh period: In the data statistics analysis, the main controller and

the cloud partition balancers need to refresh the information at a fixed period. If the

period is too short, the high frequency will influence the system performance. If the

period is too long, the information will be too old to make good decision. Thus, tests

and statistical tools are needed to set reasonable refresh periods.

A better load status evaluation: A good algorithm is needed to set Load degree high

and Load and the evaluation mechanism needs to be more comprehensive.

28



Find other load balance strategy: Other load balance strategies may provide better

results, so tests are needed to compare different strategies. Many tests are needed to

guarantee system availability and efficiency.

BIBLIOGRAPHY:

Appendix A List of Useful Websites

http://www.ijiset.com/v1s5/IJISET_V1_I5_63.pdf

http://www.ijircce.com/upload/2014/february/4_Load.pdf

http://www.iosrjournals.org/iosr-jce/papers/Vol16- issue1/Version-

6/O016168287.pdf

http://www.ijarcsse.com/docs/papers/Volume_4/3_March2014/V4I3-0303.pdf

http://www.ijetae.com/files/Volume4Issue7/IJETAE_0714_117.pdf

Appendix B List of Useful Book

Cloud Computing Bible By Barrie Sosinsky

J2EE: The Complete Reference by James Edward Keogh

Learning PHP, MySQL & JavaScript: With JQuery, CSS & HTML5

By Robin Nixon

MySQL Stored Procedure Programming

By Guy Harrison, Steven Feuerstein