Governors State University

OPUS Open Portal to University Scholarship

All Capstone Projects Student Capstone Projects

Spring 2016

Dynamic Workload Management in Hybrid Cloud

Computing
Mohid Emrooz
Governors State University

Mohammad Munawar Khan

Governors State University

Qutub Mohiuddin
Governors State University

Follow this and additional works at: http://opus.govst.edu/capstones

Part of the Computer Sciences Commons

Recommended Citation
Emrooz, Mohid; Khan, Mohammad Munawar; and Mohiuddin, Qutub, "Dynamic Workload Management in Hybrid Cloud
Computing" (2016). All Capstone Projects. 209.
http://opus.govst.edu/capstones/209

For more information about the academic degree, extended learning, and certificate programs of Governors State University, go to
http://www.govst.edu/Academics/Degree_Programs_and_Certifications/

Visit the Governors State Computer Science Department

This Project Summary is brought to you for free and open access by the Student Capstone Projects at OPUS Open Portal to University Scholarship. It
has been accepted for inclusion in All Capstone Projects by an authorized administrator of OPUS Open Portal to University Scholarship. For more
information, please contact [email protected].
 2016

Table of Content
1 Feature Description ............................................................................................................................... 3
1.1 Competitive Information ................................................................................................................. 3
1.2 Relationship to Other Applications ................................................................................................. 4
1.3 Future Enhancements ...................................................................................................................... 8
1.4 Definitions and Acronyms .............................................................................................................. 8
2 Technical Description ............................................................................................................................. 9
2.1 Project/Application Architecture..................................................................................................... 9
2.2 Project/Application modules ......................................................................................................... 10
2.3 Interactions with other Projects (if Any) ....................................................................................... 14
2.4 Risk Assessment and Management ............................................................................................... 14
3 Project Requirements ............................................................................................................................ 15
4 Project Design Description ................................................................................................................... 17
5 Project Design Units Impacts ............................................................................................................... 26
6 Open Issues ........................................................................................................................................... 26
7 Acknowledgements ................................................................................................................................ 27
8 References ............................................................................................................................................. 28

2
 1. Description

1.1 Competitive Information

The main core part of this hybrid computing model, a very intelligent or skilled workload
managing service has been designed for better throughput of proactive workload management. It
allows a collaboration in between on- and off-premise clouds structures for hosting Internet-
based applications for clients in the organization, also the its main striking feature lies in the
explicit segregation of its modules like base workload and flash crowd workload, The main
benefit of this application is the that its service uses a very fast frequent data item detection
algorithm, which allows the application to be a robust and very fast processing of large volume
of data.
In this project we are going to demonstrate that off premise cloud is much better than a wide
range of utilization administrations and particularly we can demonstrate it for both open and
private mists. On reason implies on area, though off premise implies remote (in the cloud). Case
in point if an application keeps running on an "on-reason" server it implies the server is
physically in the organization. In the event that you have an off reason arrangement it's
facilitating in the cloud or unified area.

3
1.2 Relationship to other applications:

It is much focuses on stand out of the solicitations taking care of limit of either open cloud or
private cloud approach yet it is prescribed to take both of the mists under one server as it were.
VMware additionally proposed in its Virtual Data Centre Operation System outline the V- cloud
administration idea which helps endeavor clients grow their inside IT base into an inward cloud
model or influence off-reason registering limit. At the point when current IT frameworks
advance from the devoted stage model to the common stage model along the distributed
computing pattern, we trust a center innovation segment in need is an adaptable workload
administration plan working for both models, and our workload considering innovation is
proposed as one response for it

Applications: The scalability of the application is often based on how well it can be applied in
the real time work .this applications proves to be an acceptable properties which can be
utilized in a variety of data concepts .this behavior of the application is amazingly
striking feature

Some of the areas that this application can be used can be hybrid cloud deployment services and
support. A hybrid cloud is an integrated cloud service which can utilize private and public clouds
to carry out certain operations and allow access to local resources for clients that are connected
within the organization. All cloud computing should support such tasks but public cloud are
more cost saving when compared to private cloud

4
Some of the other applications case study that are connected with our applications:

Title : Finding the Frequent Items in Streams of Data

Author : Graham Cormode, Marios Hadjieleftheriou

Year : 2008

Description: The successive things issue is to prepare a flood of things and discover every one of
those which happen more than a given part of the time. It is a standout amongst the most
vigorously considered issues in mining information streams, going back to the 1980s. Numerous
different applications depend specifically or in a roundabout way on finding the regular things,
and executions are being used in expansive scale modern frameworks. In this paper, we depict
the most imperative calculations for this issue in a typical structure. We put the distinctive
arrangements in their chronicled setting, and depict the associations between them, with the point
of clearing up a portion of the disarray that has encompassed their properties. To facilitate
delineate the distinctive properties of the calculations, we give standard usage. This permits us to
give observational confirmation that there is significant variety in the execution of incessant
thing calculations. As well as can be expected be actualized to discover continuous things with
high precision utilizing just several kilobytes of memory, at rates of a large number of things
every second on shabby advanced equipment.

5
Title : Memory Resource Management in VMware ESX Server

Author : Carl A. Waldspurger

Year : 2002
Description: VMware ESX Server is a dainty programming layer intended to multiplex
equipment assets effectively among virtual machines running unmodified ware working
frameworks. This paper presents a few novel ESX Server components and strategies for
overseeing memory. A swelling strategy recovers the pages considered minimum significant
by the working framework running in a virtual machine. An unmoving memory charge
accomplishes proficient memory use while keeping up execution separation ensures.
Content-based page sharing and hot I/O page remapping abuse straightforward page
remapping to take out repetition and diminish replicating overheads. These systems are
consolidated to proficiently bolster virtual machine workloads that over submit memory.

Title : Multilevel k-way Hyper graph Partitioning.

Author : George Karypis, Vipin Kumar.
Year : 1999

Description: Hyper diagram dividing is an imperative issue with extensive application to

numerous territories, including VLSI outline, productive capacity of expansive databases on
circles, and information mining. The issue is to parcel the vertices of a hyper chart into k
generally equivalent amounts of, such that a specific target capacity outlined over the hyper
edges is upgraded. A normally utilized target capacity is to minimize the quantity of hyper
edges that traverse distinctive parcels; nonetheless, various other target capacities are
likewise viewed as valuable. The most normally utilized methodology for figuring a k-way
dividing depends on the recursive separation world view that diminishes the issue of
6
registering a k-route apportioning to that of per-framing an arrangement of bisections. The
issue of processing an ideal separation of a hyper diagram is NP-hard; nonetheless, numerous
heuristic calculations have been produced. The overview by Alpert and Kahn gives a point
by point portrayal and examination of different such plans.
Title : Intelligent Workload Factoring for A Hybrid Cloud Computing Model
Author : Hui Zhang, Guofei Jiang, Kenji Yoshihira
Year : 2008
Description : Despite the build-up, distributed computing still remains focused
directors' watching list. The preventions to the selection of open distributed computing
administrations into big business IT incorporate administration unwavering quality, security
and information protection, control consistent prerequisites, etc. To address those worries, we
propose a cross breed distributed computing model which endeavour clients may embrace as
a suitable and cost-sparing procedure to make the best utilization of open cloud
administrations alongside their exclusive (legacy) server farms. An insightful workload
figuring administration is intended for this half and half distributed computing model. It
empowers league amongst on-and off-reason bases for facilitating an Internet-based
application, and the knowledge lies in the express isolation of base workload and trespassing
workload, the two normally distinctive parts in application workload. The centre innovation
of the insightful workload considering administration is a quick continuous information thing
recognition calculation, which empowers figuring approaching solicitations on volume as
well as on information content, after changing application information prevalence.
Through investigation and broad assessment with genuine follow driven reproductions and
analyses on a half and half testbed comprising of nearby register stage and Amazon Cloud
administration stage [1], we demonstrated that the workload figuring innovation can
empower solid workload forecast in the base burden zone (with straightforward factual
system), accomplish asset productivity (e.g., 78% higher server limit than that in base burden
zone) and decrease information store/replication overhead (up to two requests of greatness)
in the trespassing load zone, and respond quick (with a X2 speed-up element) to the changing
application information notoriety upon the landing of burden spikes.

7
8
1.3 Future Enhancement:
For future enhancement’s there has been many algorithms that sound very appropriate for
improvement in the quick access to the frequent data files or record which could improve overall
performance of the applications and also better fault less algorithm are taken in to consideration
in order to increase the reliability of the applications
1.4 Definitions and acronyms:
Some of the definitions and famous acronyms in this applications are
Maximal intelligent subset recognition

Definition:

As various gatherings of criteria can give clashing data about the information unwavering
quality, MCS shows up as a characteristic answer for unravel these contentions while protecting
a maximal measure of data.

1. Infrastructure-as-a-Service (Iaas),
2. Platform-as-a-Service (Paas) (GoogleApps are examples of PaaS), &
3. Software-as-a-Service (Saas)

9
 2.Technical Description

2.1 Project/Application Architecture

In this segment, we show the sort of data we have considered to assess the unwavering
quality of trial information in Life Science. These criteria are components that are typically
found in distributions reporting test results. Note that the greater part of these criteria are not
particular to Life Sciences, and can be utilized for any exploratory information. The rundown
of criteria is, obviously, not thorough. For other prevalent cases, for example, touristic
information or different utilizations of the Semantic Web, a few criteria utilized here are all
sufficiently inclusive to be legitimate, yet they should be finished by other appropriate
criteria.

2.1 Modules
This project having the following six modules

 Self Service Authentication Access.

 Multi Remark IT Control.
 Intelligent Work Load Factoring.
 Flash Crowd Load Infrastructure.
 Standardized Service Delivery.

10
Modules Description:

Self-service authentication access

In this module we outline the windows for the undertaking. These windows are utilized to
communicate something specific starting with one companion then onto the next. We utilize the
Swing bundle accessible in Java to plan the User Interface. Swing is an appliance toolbox for
Java. It is a piece of Sun Microsystems' Java Foundation Classes an API for giving a graphical
client interface to Java programs. In this module chiefly we are centering the login plan page
with the Partial learning data. Application Users need to see the application they have to login
through the User Interface GUI is the media to associate User and Media Database and login
screen where client can enter his/her client name, secret word and watchword will check in
database, if that will be a substantial username and secret word then he/she can get to the
database.

Multi Remark IT Control

Distributed computing suppliers like Vendors through our web2office stage convey
applications by means of the web, which are gotten to from web programs and desktop and
versatile applications, while the business programming and information are put away on servers
at a remote area. At times, legacy applications (line of business applications that as of not long
ago have been predominant in dainty customer Windows registering) are conveyed by means of
a screen-sharing innovation, while the processing assets are solidified at a remote server farm
area. At the time sharing assets it will control a wide range of in regards to the off reason cloud
whether it is open or private mists without having any deviations.

11
Intelligent Work Load Factoring

At the season of facilitating itself we can take extremely mind towards the
arrangement of use. Clever work load calculating itself saying that how effectively we
can deal with the information of the server. To lessen the server workload later, on the off
chance that you consider the current instruments there dependably want to store a wide
range of cloud information in same sort of environment. In any case, in the proposed
system we absolutely resolve those sort of issue related components, By utilizing the
Intelligent Workload calculating to store one sort of information in their own particular
related environment and another sort of information we will store in their related
surroundings. By this procedure we can lessen the workload in further.

Flash Crowd Load Infrastructure

In Flash Crowd Load Infrastructure what is the significant procedure is happens is

distributing the assets to the shoppers from the suppliers through the cloud. So this
module how it satisfies implies it takes the quantity of questions from the quantity of
shoppers at once and it forms every single solicitation given by the client and how it
will recognize the solicitation implies every single inquiry is going to part as various
catchphrases and in view of that watchwords match and tally we will break down the
asset what they are going to asking. We keep up independent neighborhood store
memory for the quantity of solicitations to lessen the group burden to the purchaser
demands.

12
Standardized Service Delivery

This is the last module of our undertaking in this Monitoring of Cloud and administration
conveyance is a noteworthy assignment of principal significance for both Cloud Service
Providers and Cloud Service Consumers. On the one side, it is a key apparatus for controlling
and overseeing assets and programming frameworks; on the other side, it gives data and Key
Performance Indicators (KPI) for both stages and applications. The nonstop checking of the
Cloud and of its Service Level Agreements (SLAs), for instance, as far as accessibility,
postponement, and so forth supplies both the Providers and the shoppers data, for example, the
workload produced by the last mentioned or the execution and Quality of Service (QoS) offered
through the Cloud, likewise permitting to actualize components to counteract or recuperate
infringement, for both the Provider and Consumers. Distributed computing includes numerous
exercises for which observing is a crucial errand.

Module Diagrams:
Self Service Authentication Access

User
2 Interactions with other Projects (if Any)Error! Bookmark not defined.
Regist
3 Interactions with other ApplicationsError! Bookmark not defined.
4 CapabilitiesError!
User Bookmark not defined.
Interacti
5 Project RequirementsError! Data
on Bookmark not defined.
6 3.1 Identification of RequirementsError! Bookmark not defined.
7
User Error not defined.
8 3.3 Security and Fraud PreventionError! Bookmark
Window Page
9 3.4 Release and Transition Plan Error! Bookmark not defined.

13
Intelligent Work Load Factoring

Owner Applying
L i C d ti l Data Base

Error

Hybrid Valid
Cloud invalid

Work Load Factoring with

Multiple Managers
Mana Mana
ger1 ger2

Flash Crowd Load Infrastructure

User Applying Data

User Query

Crowd Load
Effi i t SValidh
Keywords
invalid
R i
14
Standardized Service Delivery

User

User Login Interaction

Data Base

User Window
Monitoring Querying Message

R
R 2

Resource n

2.3interactions with other projects:

At this current situation we have not compared our work with any other projects that are being
running in the same platform

2.4 risk assumptions:

The risks that has to consider in this application as this is a fast moving era and cloud computing
has been in a peak since a decade new applications can be a risk to this application.

15
 3. Project Requirements

REQUIREMENTS

The equipment prerequisites may serve as the premise for an agreement for the application of the
framework and ought to be a finished and reliable particular of the whole system. They are
utilized by programming engineers as the starting stage for the framework outline. It shows what
the framework do and not how it ought to be implemented.

Gear

PROCESSOR : PENTIUM dual core

RAM : 4 GB

MONITOR : 15" Shading

HARD Circle : 200 GB

The product prerequisites report is the detail of the framework. It ought to incorporate both a
definition and a determination of necessities. It is proper set of what the framework ought to do
as opposed to how it ought to do it. The product prerequisites give a premise to making the
product necessities detail. It is helpful in cost evaluating, arranging exercises of group,
performing assignments and following the groups and it's advancement all through the
betterment action.

Front End : (JSP, SERVLET), STRUTS

Back End : MS SQL 2000/05

Operating system : Windows 08

IDE : Net Beans, Shroud

16
Utilitarian Prerequisites An utilitarian prerequisite compromises an element of a product
framework or its part. A capacity is depicted as an arrangement of inputs, the conduct, and
yields. The proposed framework is accomplished by Booking Process.

NON-Utilitarian Prerequisites

Productivity

Utilizing the calculation like the Planning technic to secure proficiently all the matter that are
utilized as a part of the provided application.

17
 4. Project design description:

Plan Building manages the different UML [Unified Displaying language] outlines for the
execution of venture. Outline is an important building representation of a thing that is to be
constructed. Programming outline is a procedure through which the prerequisites are deciphered
into representation of the product. Outline is the spot where quality is rendered in programming
building. Outline is the way to precisely make an interpretation of client necessities into
completed item.

Use Case Diagram:

EXPLANATION:

The huge reason behind an utilization case outline is to show what framework breaking points
are performed for which on-screen character. Parts of the performing masters in the framework
can be portrayed. In our utilization case diagram first client login into client window then on the
off chance that it is a liberal client proposes then it can chat with the cloud server. The bona fide
extraordinary position of the thing is if customer asks for a few information then the deals will
reach to the server then the information which is asked for by the customer is there with the
cloud server then it will be dealt with and practically sending information to the asked for
customer. On the off chance that the information is not there with the cloud server then it needs
to look in its related customers and if the point of interest is there means then it will be shared to

18
the specific requester.

Class Diagram:

EXPLANATION:

. In our class graph we having the insights about client, on the off chance that it is a legitimate
client implies then it can speak with the cloud server. The real point of interest of the thing is if
customer asks for a few information then the solicitation will reach to the server then the
information which is asked for by the customer is there with the cloud server then it will be
handled and successfully sending information to the asked for customer. In the event that the
information is not there with the cloud server then it needs to seek in its related customers and if
19
the asset is there means then it will be shared to the specific requester.

OBJECT DIAGRAM:

client
client

Server

client

EXPLANATION:

to start with client login into client window then in the event that it is a substantial client implies
then it can speak with the cloud server. The significant point of preference of the thing is if
customer asks for a few information then the solicitation will reach to the server then the
information which is asked for by the customer is there with the cloud server then it will be
prepared and successfully sending information to the asked for customer. On the off chance that
the information is not there with the cloud server then it needs to look in its related customers
and if the asset is there means then it will be shared to the specific requester.

20
State Chart:

EXPLANATION:

The real point of preference of the thing is if customer asks for a few information then the
solicitation will reach to the server then the information which is asked for by the customer is
there with the cloud server then it will be prepared and viably sending information to the asked
for customer. On the off chance that the information is not there with the cloud server then it
needs to seek in its related customers and if the asset is there means then it will be shared to the
specific requester.

21
Activity Diagram:

EXPLANATION:

In any case customer login into customer window then if it is a honest to goodness customer
suggests then it can talk with the cloud server. The noteworthy purpose of inclination of the thing
is if client requests a couple data then the sales will go to the server then the data which is
requested by the client is there with the cloud server then it will be readied and satisfactorily
sending data to the requested client. In case the data is not there with the cloud server then it
needs to look in its related clients and if the benefit is there means then it will be shared to the
particular requester.

22
Sequence Diagram:

EXPLANATION:

In our social occasion outline exhibiting outlines work with each other and all together. In this
first client login into client window then in the event that it is a genuine client proposes then it
can converse with the cloud server. The certifiable motivation behind energy of the thing is if
customer asks for several information then the deals will reach to the server then the information
which is asked for by the customer is there with the cloud server then it will be prepared and
enough sending information to the asked for customer. In the event that the information is not
there with the cloud server then it needs to search for in its related customers and if the
advantage is there means then it will be shared to the specific requester.

23
Collaboration Diagram:

EXPLANATION:

A coordinated effort outline depicts collaborations among items regarding sequenced messages.
Joint effort outlines speak to a mix of data taken from class, arrangement, and use case graphs
depicting both the static structure and element conduct of a framework. In this a client login into
client window then in the event that it is a legitimate client implies then it can speak with the
cloud server. The real point of interest of the thing is if customer asks for a few information then
the solicitation will reach to the server then the information which is asked for by the customer is
there with the cloud server then it will be handled and viably sending information to the asked
for customer. In the event that the information is not there with the cloud server then it needs to
seek in its related customers and if the asset is there means then it will be shared to the specific
24
requester.

DATA FLOW DIAGRAM:

LEVEL-0:

Client login

Error Page
Data Base

Client Window

LEVEL-1

Client window

Implement
Cloud Resource ation of
Management O ti l
Services

FeedBack Unit
25
EXPLANATION:

It doesn't demonstrate data about the planning of procedures, or data about whether procedures
will work in arrangement or in parallel. In the DFDs the primary client login into client window
then in the event that it is a substantial client implies then it can speak with the cloud server. The
significant favorable position of the thing is if customer asks for a few information then the
solicitation will reach to the server then the information which is asked for by the customer is
there with the cloud server then it will be prepared and successfully sending information to the
asked for customer. On the off chance that the information is not there with the cloud server then
it needs to seek in its related customers and if the asset is there means then it will be shared to the
specific requester.

Component Diagram:

EXPLANATION:

In the United Exhibiting Tongue, a portion diagram portrays how sections are wired together to
shape greater parts and they are used to plot the structure of subjectively complex systems. For
this in our portion chart we did the essential customer login into customer window then if it is an
authentic customer suggests then it can talk with the cloud server. The genuine great position of
the thing is if client requests a couple data then the sales will reach to the server then the data

26
which is appealed by the client is there with the cloud server then it will be readied and feasibly
sending data to the requested client. In case the data is not there with the cloud server then it
needs to look for in its related clients and if the benefit is there means then it will be shared to the
particular requester

5. Project Design Units Impacts:

Our application does not have any design impacts at this period.

6. Open issues

As the application is based on cloud computing and there are hundreds of applications that are
being currently running of the cloud computing technology therefore the durability of the
applications success can be a issue as the new technologies in the cloud area is been known to
increase in numbers since last decade and it is expected to become more popular in the coming
time.so their can a possibility of similar applications been created which can lower the value of
this application.

27
 7. Acknowledgment:

This project was made successful under the guidance of professor DR.DO YOUNG PARK-
governors state university. He had helped us with providing his useful guidance and tips which
made our work possible. We would also like to show our gratitude to Professor Nelson Chen and
Professor Dr soon-park for providing us with valuable information which helped us to a great
extent in completion of this project.

28
REFERENCES
1. GigaSpaces. (n.d.). Retrieved May 06, 2016, from http://www.gigaspaces.com/

2. Hello Triby - Meet Alexa. (n.d.). Retrieved May 06, 2016, from
https://developer.amazon.com/

3. Kang, X., Zhang, H., Jiang, G., Chen, H., Meng, X., & Yoshihira, K. (2008). Measurement,
Modeling, and Analysis of Internet Video Sharing Site Workload: A Case Study. 2008 IEEE
International Conference on Web Services. doi:10.1109/icws.2008.28

4. Content Delivery Network. (n.d.). Retrieved May 06, 2016, from

http://www.level3.com/en/products/content-delivery-network/

29