Optimal Multi-server

Configuration for Profit

Maximization in Cloud Computing
Presentation by:
E.SOWMYA(15S11A0550)
B.SRIRAM(15S11A0551)
K.V.RAKESH(15S11A0555)

Under the guidance of:S.Md.Mujeeb

Overview

 Abstract
 Existing system with disadvantages
 Proposed system with advantages
 Requirements
 Architecture
 UML diagrams
 Modules
CSE DEPARTMENT-MRITS 2
 In this project, we take customer satisfaction into

consideration to address the development of cloud

Abstract computing and an increasing number of enterprises start to

adopt cloud service.

Customer satisfaction affects the profit of cloud service

providers in two ways:

1.The cloud configuration affects the quality of service

which is an important factor affecting customer

satisfaction.

2.The customer satisfaction affects the request arrival

rate of a cloud service provider.

3 CSE DEPARTMENT-MRITS
 The cloud service providers does not care about
customer satisfaction in proper direction.

Existing
system
The cloud configuration affects the quality of service
which is an important factor affecting customer
satisfaction.

The customer in return expects fast request arrival

rate which is not provided by the cloud service
providers.

However, the existing works considering customer

satisfaction do not give a proper formalized definition
for it which gives the result of the profit is low

CSE DEPARTMENT-MRITS
4
 Customer satisfaction does not
give proper direction for it.
Disadvantages
Quality of service very poor.

Price of service is high.

The results of profit were very

poor.

CSE DEPARTMENT-MRITS 5
 1.In this system, we take customer satisfaction into
consideration to address this problem.

Proposed
2.We firstly refer to the definition of customer
system satisfaction in economics and develop a formula for
measuring customer satisfaction in cloud computing.

3.Then, a optimal multi server configuration is used to

attract more customers by providing customer friendly
environment.

4.Lastly, taking into consideration customer satisfaction,

service-level agreement, renting price, energy
consumption and so forth, a profit maximization problem
is formulated and solved to get the optimal configuration
CSE DEPARTMENT-MRITS
such that the profit is maximized. 6
 Customer satisfaction was addressed in
proper direction.

Advantages Quality of service is increased.

The request arrival rate is fast.

The price of cloud service is reduced.

Profit is maximized.

CSE DEPARTMENT-MRITS
7
 Hardware Requirements

Requirements
System : Pentium IV 2.4 GHz.

Hard Disk : 40 GB.

Floppy Drive : 1.44 Mb.

Monitor : 15 VGA Colour.

Mouse : Logitech.

Ram : 512 Mb.

CSE DEPARTMENT-MRITS 8
 Software Requirements

Operating system : Windows XP/7.

Coding Language : JAVA/J2EE

IDE : Netbeans 7.4

Database : MYSQL

CSE DEPARTMENT-MRITS 9
 Architecture

CSE DEPARTMENT-MRITS
10
 1.A CUSTOMER is the receiving end of a service or the
requestor of a service in a client/server model type of system. The
client is most often located on another system or computer, which
can be accessed via a network. This term was first used for
Architecture devices that could not run their own programs, and were
connected to remote computers that could via a network.

2.A cloud SERVICE PROVIDER, or CSP, is a company that

offers some component of cloud computing -- typically
infrastructure as a service (IaaS), software as a service (SaaS) or
platform as a service (PaaS) -- to other businesses or individuals

3.In general, cloud service providers make their offerings available

as an on-demand, self-provisioning purchase. Additionally,
customers can pay for the cloud-based services on a subscription
CSE DEPARTMENT-MRITS basis -- for example, under a monthly or
11 quarterly billing structure.
 Register

Cloud Request
UML Diagrams
Uploaddata
Admin

Fig.01: forwaord admin

Use case diagram Customer

accept req

view users

Intermediate(Broker)

rental space

Bae, Rrenta response

CSE DEPARTMENT-MRITS 12
 Customer Intermediate(Broker) Cloud Admin CloudSpace

1 : Register()

Fig.02: 2 : Send Request()

3 : forward to admin()
Sequence diagram
4 : req accepted()

5 : upload files()

6 : used memory()

7 : rental memory request()

8 : rentamemory accept()

9 : rental space view()

CSE DEPARTMENT-MRITS
13
 Fig.3:
Class diagram

Intermediate(Broker)
+String loginid
+String pswd
+acceptRequest()
Customer
+viewUsers()
Cloud Admin
+String cname
+String pswd +String loginName
+int id +String pswd
+long memorysize
+acceptRequest()
+customerRegister() +viewusers()
+uploadData() +memoryManagement()
+availabaleSpace()
+usedSpace()
+rentalSapce()
Cloud Space
+float memoingb

+totalspace()
+usedSpace()
+availablespace()
+registercustomer()

Cse department mrits 14

 Fig.4:
Activity diagram
Intermediate(Broker)
Customer Cloud Admin

Send Cloud Space Request

View Request View Cust Request

Upload Data

Forward to Admin Accept Memory Request

Memory Space

View users View Users

Rental Request

Rental Space

Cse department mrits 15

 1.Cloud Service Model

Modules 2.Service-Level
Agreement

3.Customer Satisfaction
Aware Arrival Rate

4.The Cost Model

CSE DEPARTMENT-MRITS 16
  The profit optimization of cloud service providers, we should
study how to measure the customer satisfaction of a cloud
service provider and how the customer satisfaction affects the

Module 1: 
its profit, which are analyzed in the following.
In Cloud Server the M/M/m model, m is the number of servers,

cloud service
and all servers run at an identical speed s (measured by the
number of instructions that can be executed in one unit of time).
Assume that the interarrival times of service requests are

model independent and identically distributed exponential random

variables, in other words, the arrival requests follow a Poisson
process with arrival rate λ. The execution requirements of the
tasks (measured by the number of instructions to be executed)
are exponential random variables r with mean r.
 Since the server execution peed is s, the service times of the
requests are also exponential random variables x = r/s with
mean x = r/s. Hence, the average service rate, i.e., the average
number of service requests that can be completed by a server
with speed s in one unit of time, is μ = 1/x = s/r.

CSE DEPARTMENT-MRITS
17
  In general, the QoS is affected by many factors such
as the service time, the failure rate and so forth.

Module 2: However, in this paper, we measure the QoS of a

request by its response time for two reasons.

Service-level 1. First, the service time is easily measured.

agreement 2. Second, it gives customers an intuitive feeling of

QoS. For customers, they do not care how failures are
managed when failures occur. They only care
whether the task can be completed successfully and
how long it takes.

CSE DEPARTMENT-MRITS
18
  Assume that the total market demand is λmax, the
market share MS of a service provider is the ratio of
the actual task arrival rate λ and λmax which can be
Module 3: formulated as MS = λ/λmax.

customer  In general, a higher customer satisfaction would

lead to a larger market share, but the growth trends

satisfaction are different in different situations. In this paper, we

assume that the market share MS of a service

Aware Arrival provider is linearly increasing with its customer

satisfaction which is denoted as MS = S.

Rate  Combining above two equations, we can get the

relationship between the actual task arrival rate λ
and the custom satisfactionMS as λ = Sλmax.

CSE DEPARTMENT-MRITS
19
  The cost of a service provider is mainly used to pay
the rent and the electricity fee. A service provider
rents servers from an infrastructure provider and
pays the corresponding rent.
 The rent is determined by the number of rented

Module 4: servers and the rental price per server per unit of
time.

The cost model  Assume that the rental price of one server per unit
of time is β, and m servers are rented. The rent per
unit of time is calculated as Erent = βm.

CSE DEPARTMENT-MRITS
20
21
Scope of the project
 The cloud computing has become a booming service industry and in recent years
the cloud service providers increase in large numbers.

 There are some works in cloud computing which consider customer satisfaction.

 The cloud service provider operating geographically distributed data centers in a

multi-electricity-market environment.

22

CSE DEPARTMENT-MRITS
23