Contenu connexe Similaire à A Virtualization Model for Cloud Computing (20) A Virtualization Model for Cloud Computing1. Full Paper
©2012ACEEE
DOI:02.ACS.2012.3.
Proc. of Int. Conf. on Advances in Computer Science 2012
512
AVirtualization ModelforCloud Computing
Souvik Pal1
,Suneeta Mohanty2
, P. K. Pattnaik3
and G.B.Mund4
Email: souvikpal22@gmail.com
1,2,3,4
KIIT University, Bhubaneswar, India
Email: {smohantyfcs@kiit.ac.in, patnaikprasantfcs@kiit.ac.in,mund@kiit.ac.in}
Abstract—Cloud Computing is now a very emerging field in
the IT industry as well as research field. The advancement of
Cloud Computing came up due to fast-growing usage of
internet among the people. Cloud Computing is basically on-
demand network access to a collection of physical resources
which can be provisioned according to the need of cloud user
under the supervision of Cloud Service providerinteraction.
From business prospective, the viable achievements of Cloud
Computing and recent developments in Grid computing have
brought the platform that has introduced virtualization
technology into the era of high performance
computing.Virtualization technology is widely applied to
modern data center for cloud computing. Virtualization is
used computer resources to imitate other computer resources
or whole computers. This paper provides a Virtualization
model for cloud computing that may lead to faster access and
better performance. This model may help to combine self-
service capabilities and ready-to-use facilities for computing
resources.
Index Terms—Cloud Computing, Virtual Machine,
Virtualization, Hypervisor.
I. INTRODUCTION
Cloud Computing or Internet Computing is a model or
infrastructure that delivers applications as services over the
Internet and also provides services that involve hardware
and system software in the data centers [1]. Cloud is a type
of distributed and parallel system comprising of a collection
of interconnected and virtualized computers which are
dynamicallyprovisioned and presented as one or more unified
computing resources based on service-level agreement [2].
For simplicity, A Cloud itselfan infrastructure or framework
that comprises a pool of physical computing resources i.e. a
set of hardware, processors, memory, storage, networks and
bandwidth, which can be organized on Demand into services
that can grow or shrink in real-time scenario[3].
In this era of immense usage of internet throughout the
globe, virtualization technology is the key feature of Cloud
Computing. Virtualization technologycreatesan environment
that enables on-demand and convenient network access to a
shared collection of configurable physical resources (i.e. set
of hardware, processors, memory, storage and bandwidth)
and as well as helps the creation of individual Virtual
Machines (VM) according to the need of the cloud user. This
paper leads to a Virtualization model which elaborates the
Virtualization layer, job deployment onto the VMs, and the
interconnection between cloud user, hypervisor and Virtual
Machines.
A. Brief History
In this era ofrapidlygrowing usage ofinternet throughout
the world, Cloud Computing has become the icon ofInternet-
centric business place in the IT industry. The Cloud
Computing is not a totally new technology; it is basically a
journey through distributed, cluster and grid computing. In
the case of cluster computing [4], many clusters were formed.
A cluster is an aggregation of homogeneous computers
interlinked with each other by high-speed networks [5] and
allocated in a centralized way.Sometimes, it is not feasible for
a single computer to execute massive compute and data
intensive tasks. When a huge computational task is needed
to execute, the individual computers may form a cluster to
share computational workload and they function as a single
virtual machine. The advantage is that failure of a node
doesn’t put anyimpact over the system because some another
standbynode will dothe needful task. But the drastic changes
in the no. of service-oriented requests, lack of sufficient
additional resources, centralized resource handling are the
issues of cluster computing. We need to discuss the updating
issues in Grid Computing [6].
Grid computing is the combination of computers from
various administrative domains tomeet a common target and
maydisappear after the completion of the task. It is analogous
topower grid [7]. In an electrical grid scenario, there are three
phases namely, Generation, Transmission and Distribution.
But the users don’t need to know either the location of power
plant or the process by which users getting the power.
Likewise, in grid computing, the end users donot knowabout
the underlying processes and resources. A massive
computing task is assigned to a grid and the grid shares out
the task into sub tasks, which are independently consigned
to a range of nodes. It’s all done by Job-scheduling Process
[8]. Like cluster computing, Failure of a node doesn’t affect
the whole process. But it’s a different task to organize and
maintain heterogeneous systems. The grid resources incur
unnecessaryoverheads in terms of stability. In compared to
cluster and grid computing, clouds are highly scalable,
capable of both centralized & distributed resource handling,
loosely coupled and provide on-demand computation &
application service. Cloud computing is basically known as
computing over internet. Cloud computing is an enhancement
ofdistributed and parallel computing, Cluster Computing and
Grid computing. In this advanced era, not only user able to
use a particular web based application but also that may be
in activeparticipation in its computational procedure byeither
adopting ,demanding or pay-per-use basis [9][10].
10
2. Full Paper
©2012ACEEE
DOI:02.ACS.2012.3.
Proc. of Int. Conf. on Advances in Computer Science 2012
512
B. Traditional versus Virtualization Approach
Virtualization is used computer resources to imitate other
computer resources or whole computers. A virtualization
environment that enables the configuration of systems (i.e.
compute power, bandwidth and storage) as well as helps the
creation of individual Virtual Machine has become the
essential technology of cloud computing environments.
Virtualization provides a platform for optimizing complex IT
resources in a scalablemanner (efficientlygrowing), which is
ideal for delivering services. At a fundamental level,
virtualization technology enables the abstraction or
decoupling of the application payload from the underlying
physical resource [2] that means Physical resources can be
changed or transformed into virtual or logical resources on-
demand which is called Provisioning.
In traditional approach, there are mixed hardware
environment, multiple management tools, frequent application
patching and updating, complex workloads and multiple
software architecture. But comparativelyin cloud data center
far better approach like homogeneous environment,
standardize management tools, minimal application patching
and updating, simple workloads and single standard software
architecture [11].
The paper organized as follows:
In the section 2, we have discussed the deployment models.
Section 3 has given the idea of conventional cloud service
models. And in the section 4, we have given a proposed
Virtualization mapping model.Section 5elaborates the
mapping approach &lastly, section 6 leads to future work.
II. CLOUD DEPLOYMENT MODELS
Depending on the requirement of services and the
applications, cloud models are categorized into four types
[12]:
Public Cloud: Public cloud is an open cloud where the
computer resources and same infrastructure used bymultiple
users. Usable standardized workload for applications, high
scalability, testing and developing applications, security
strategyare the basic criterion of public cloud. Email-system
is a good example of public cloud. For example, when end-
users use mail services say Yahoo or Gmail account, they
need to have only a computer and internet connection. User
need not to know about the underlying process. They are
just using the cloud. All the maintaining, testing and
developing task are done byYahoo or Google itself.
Private Cloud: Private cloud is a particular organization-
based Cloud where the computer resources and infrastructure
are handled by an organization and highly virtualized data-
center is located inside the client’s firewall. It has specific
workloads and provides well-managed environment,
optimized usage of computing resources, security and
compliance. Quality of Service (QoS) is managed and
controlled by the organization itself. For example, Amazon
Virtual Private Cloud (VPC) is a private cloud, which offers
clients isolated AWS (Amazon Work Space) and protection
byVirtual Private Network (VPN) connections.
Hybrid Cloud: Combination of private cloud and public
cloud makes a hybrid cloud. The vendors use the physical
resources from public cloud on the basis of need of thecloud
user and return the resources when it is of no use. Let us
consider a scenario: Suppose a company want to use a SaaS
Application, which would meet the considerations of the
company, i.e. Securityand standardize usage throughout the
company. The SaaS provider creates a private cloud for the
particular company inside their firewall so that the entire
company can use the cloud as a standard. Now they provide
the companywith a Virtual PrivateNetwork (VPN) for getting
more security [13]. Hence, we can see that. Combining the
advantages of private and public cloud, we can create a hybrid
cloud.
Community cloud: In case of joint venture application, a
same cloud infrastructure needs to be constructed and shared
by several organizations jointly, so that they can use the
same framework as well as policies, services, requirements,
applications, and concerns [14]. Communitycloud is highly
scalable and reduces cost complexity. The third-partyvendor
or any one of the vendors within the community may host
and maintain the communitycloud infrastructure.
III. CLOUD SERVICE
End-users or clients use the cloud services according to
their needs. Cloud Service Providers (CSPs) deliver the
service on-demand basis. In this section, we are going to
discuss different types of services [12]providedby the CSPs
as shown in the figure [1].
Resources as a Service (RaaS)[12]:At the bottom-most
layer of the cloud service stack, there would be a collection
of physical Resources such asstorage, servers,
bandwidth,data center space, and networks,which may be
accessed and shared bymultiple CSPs. In the current scenario,
resource virtualization is the keyfeature of Cloud computing
and fast growth of hardware and computing resource
virtualization facilitates the service providers for using RaaS
service when theyare in need of the particular service. As for
example of RaaS, Amazon S3 maythe best example.
Hypervisor Monitoring Environment (HME) [12] is like a
middleware, which persists in between IaaS and physical
resources (RaaS layer). The main functionalityof this layer is
to allocate the hardware resources and computing resources
from thebase-layer i.e. RaaSacross multipleVirtual Machines
and to monitor the accesses of Virtual machines. We will
elaborately discuss about this layer in the next section.
Infrastructure as a Service (IaaS): IaaS deals with
infrastructure on which RaaSmayget expanded itself due to
resource virtualization. IaaSprovides virtualization technology
which involves the provisioning ofinfrastructure to the cloud
users. Clients may rent infrastructure services rather than
buying those servicesto reduce the cost complexity and
resource virtualization may facilitate to get those services.
Amazon EC2 is the common example of IaaS.
11
3. Full Paper
©2012ACEEE
DOI:02.ACS.2012.3.
Proc. of Int. Conf. on Advances in Computer Science 2012
512
Fig. 1. Cloud Service Model
Platform as a Service (PaaS):PaaS has the capability to
provide independent platform having deployment capabilities
and it may be capable of executing multiple application on
single platform concurrently. The user may be free to create
his applications which can be run over the service provider’s
infrastructure. Google Apps is the example of PaaS.
Software as a Service (SaaS): SaaS provides the platform
for software usage. Clients just use the software service
without the overhead of running, maintaining and updating
the software. It helps to reduce the cost complexity because
clients need not to be concerned about all those issues.
Salesforce.com can be example of SaaS.
IV. PROPOSED VIRTUALIZATION MAPPING MODEL
In the paper, we have proposed an enhanced service model
which involves in mapping ofVirtual machines onto the host
machines. While accessing operating system, network, data
storage or software applications through the Internet from
the service provider, cloud users have to send request
according to their needs and the hardware and computing
resources are provisioned and VMs are mapped by the
hypervisor administrator as shown in the figure [2].
A. Distributed Resources of Physical Hosts:
At the base level of the cloud stack, there would be a
pool of resources of physical hosts in the distributed manner.
Physical hosts is nothing but a collection of processors,
memory, CPU Cores, storage which are provisioned to the
user and logically presented as a several numbers of VMs
through the process of resource virtualization.
B. Hypervisor Monitoring Environment (HME):
Over the physical hosts, there are Hypervisor Monitoring
Environment layer that consists of hypervisor, resource
virtualization, VMs and hypervisor administrator. We have
considered a service scenario to increase the
understandability of this particular layer. Suppose a cloud
user needs an infrastructure-based service. While requesting
for infrastructure and resources, hypervisor is available from
the booting time of the system to manage the allocation of
physical and computing resources from the resource later
across multiple VMs which would be mapped depending on
the availability of the physical hosts.
Hypervisor:Hypervisor is basicallya mainframeoperating
system which allows other operating systems to run on the
same system concurrently [12]. Accesses ofVMs are controls
by the hypervisor. Hypervisors may be categorized in three
types [13]:
Native Hypervisor: These types of hypervisor are
designed to reside directly over the hardware platform for
providing better performance.
Embedded or Bare Metal Hypervisor:To get high
performance and to reduce the time complexity, embedded
hypervisors are integrated with processors on a separate
chip.
Host Hypervisor: These kinds of hypervisors act as
separate software layer over both operating system and
hardware to get improved performance.
Virtualization Layer:One of the most important layers of
cloudstack is Virtualization which mainlydeals with network,
storage and computation virtualization. The basic concept
ofVirtualization is tohideunderlying infrastructure bycreating
a logical interface. Resources are virtualized and logically
represented in this layer.
Network Virtualization: Network Virtualization involves
in implementing virtual networks within the physical server
for switching between all the virtual servers instead of multi-
pathed or multiplexed network channels by directlytrunking
them toWAN transport and multiple HBAs and NICs maybe
needed for each application provided with a high speed single
Ethernet connection [3].
Storage Virtualization: The key driver of storage
virtualization, storage networking and server virtualization
was Fibre Channel (FC) and FC-basedStorageArea Networks
(FC-SAN) which facilitated storage connectivity (veryhigh
speed) and storage solutions like point-to-point replication
and server-less back-up [3]. It also helps to optimize the
performance of the servers onwhich multiple applications
are running and this technique has enabled numerous
advancements in the datacenter.
Computation Virtualization: Computation Virtualization
leads to virtualization of computing resources. Computing
resources like server Virtualization and operating system
Virtualization has enabled the transformation from the server-
centric traditional computing to network and Internet-centric
computing. Operating system and server Virtualization create
the virtual (logical) servers which are free from constraints of
the underlying physical location and physical infrastructure
and the virtualization may facilitate to move the workloads
from source VM instance to target VM instance in real-time
which is calledLive Migration [3][15].
12
5. Full Paper
©2012ACEEE
DOI:02.ACS.2012.3.
Proc. of Int. Conf. on Advances in Computer Science 2012
512
Virtual Machine (VM): Virtual Machines are like the interface
to the user, which have own operating system (Guest OS)
hosted by the host OS. Instead of interacting with a single
computer, Virtual Machines should have the capability to
facilitate the aggregation of system and computing resources
from multiple machines and theyshould present a consistent
and unified view to the users and the applications. The basic
responsibility of VMs would beto support multiple OS and
application instances and to provide greater scalability and
better performance.
Hypervisor Administrator:In a Cloud Computing
Environment, hypervisor administrator may have the main
responsibility to control over the accesses of VMs. It should
monitor how the computing and system resources are
distributed and circulated for Virtualization and howVirtual
machines are mapped to the host machines so that the data
and computing resources can easily be retrieved from the
resource pool by the CSPs or the cloud user. In the booting
time of the system, hypervisor is accessible and hypervisor
administrator may regulate the allocation of system and
computing resources and VMs are mapped to the host
machine according to need of the user or the
application.Hypervisor administrator should have the
capability to manage both the resources within the server
(locally) and as well as the resources located in the other
servers connected to the network [3].
C. Platform Service:
Platform service should have ability to create FCAPS-
oriented (Fault, Configuration, Account, Performance, and
Security) virtual servers [3] which should allow to host the
guest operating systems and to execute the user-applications.
Integration-Service Platform: This block can have the
responsibility to integrate the resources for multiple
applications and for the guest operating system. There would
be different kinds of guest operating systems; hence this
should be platform independent.
Platform and OS Usage and Services: In that block,
virtual servers may load and host the preferred choice of
operating systems that allow the loading and executing the
user-application. In a cloud platform, application developers
create and deploy applications and do not necessarily need
to know the underlying resources or infrastructure.
Performance Measurements & Update
Configuration:Virtual servers can have the responsibilityto
manage performance measurements, performance
optimization, fault management andreliabilityandaccounting.
This block allows the service developers to make self-
managed and self-configuring business workflows.
D. Software Service:
This layer mayprovide ready-to-use software service and
the clients need not to take any overhead for buying and
maintaining the software [12]. Cloud users need not to know
the underlying architectureof cloud infrastructure; they just
use and access the software. Basic blocks of software services
mayfacilitate the creation, deliveryand usage of the softwares
which are discussed below [3].
Services Creation Platform: This block may provide
the tools which can be used for creation of applications.
Applications are termed as collection of services and they
can be created and distributed over the virtual servers created
and controlled by the Platform service.
Services Delivery Platform: Services deliveryplatform
can necessarily defined as a workflow engine that should be
capable of execution of the application and it should be able
tomanage the orchestration ofmultiple distributable workflow
elements.
Legacy Integration Services: This block may provide
theservices which can support integration of legacyor existing
application. Support for inheritance or legacy of previous
application is the key responsibility of that block.
E. Broker Service:
Cloud brokers are like the agents between the virtual
environment and business service layer. The main goal of
the cloud broker is to make optimal arrangement of virtual
environment intoCSPs which is configured and managed by
the service manager. In count to serving local requests, cloud
broker would have the responsibility for the federation of
remote sites [16]. Like the Virtualization layer, there are
network brokers, storage broker, and computation broker and
they have their each responsibility to look after the
Virtualization respectively. They should be able to move
throughout the virtual environmentand alsoremote sites until
the configuration and arrangement is satisfied. In addition,
cloud brokers has
F. Business Service:
Business service layer provides different aspects of
business controls and conditions like Business service
management, Identitycreation and validation,Authentication
and Access Control Management, Service level agreement,
and Renewal service management.
Business Service Management:This block may help to
manage the service orchestration, service conditions, and
process management. Service manager should be able to
derive a collection ofdesired resources and their configuration,
and also placement constraints according to licensing, cost,
confidentialityetc. Process management has the authorityto
schedule and to manage the processes. Service manager has
also the responsibility to monitor the deployed services and
to adjust their capacity(number of VM instances).
Identity Creation and Validation:Creation of user-
identity and its validation is the key responsibility of this
block. When a new user comes up in the market, he needs to
create his identity and this module may identify proper
validation for deployment of user-application.
Authentication and Access Control Management:
Authentication should have the responsibility for effective
governance and proper management of their authenticated
process. Access control management service enables policies
and rules for access control in reply to a request from a client
who is in need of resources and this block should also
14
6. Full Paper
©2012ACEEE
DOI:02.ACS.2012.3.
Proc. of Int. Conf. on Advances in Computer Science 2012
512
responsible for OS access control and network access control.
Service Level Agreement (SLA):SLA management is
key aspect of Business Service. SLA is concerned about
business-aligned rule and policies. Distributed resources like
network, storage and network are provisioned but should
not be the cause of a SLA violation of any application
executing with greater than predefined threshold line. Hence,
service agreement should be maintained to keep the rules of
an organization.
Renewal Service Management:Renewal service
management is responsible for renewal of policies, rules and
access controls. Validation and updated SLAs should also
be under the renewal service. This service basicallya smooth
interacting medium for Internet-centric business place where
both clients and CSPs could interact for their business and
IT Infrastructure.
V. A MAPPING APPROACH
In this paper, we will discuss a mappingapproach ofVirtual
Machines onto host machines depending on the availability
of the distributed resources [17].
We have defined our system as S where the set of Virtual
machines (V) are to be mapped onto the set of physical host
machines (H); and pool of physical resources are denoted by
P.
P = {CPU cores, Memory, Storage, I/O, Bandwidth,
Networking}.
According to the user-needs like IT infrastructure,
platform service or software usage, VM instances are created
by the hypervisor administrator who controls the mapping
ofVMs.
We have considered VS as Virtual Machine set:
VS = V1
+ V2
+ …. + Vm
= Vi
Vi
= { vc, vm, vr}
Where
vc = Number of CPU Cores
vm = Main Memory
vr = Storage Capacity
m = Number ofVirtual Machines
Now we considered HS as a Set of host machines:
HS = H1
+ H2
+…. + Hn
= Hi
Hi
= {hc, hm, hr}
Where
hc = Number of CPU Core
hm = Main Memory
hr = Storage Capacity
n = Number of host machines.
Now we divide the host set into two subsets:
HS = HSa
+ HSb
( a + b = n).
Where
HSa
= Set ofphysical machines having available resources
to host VMs and on which VMs can be mapped.
HSb
= Set of remaining physical machines not having
enough resources to host VMs and on which VMs cannot
be mapped.
Let f:Vi
HSa
be the Function which maps VM instance to
the set of physical machines having enough resources to
host the VM. There may be either one to one mapping or
many to one mapping. In one to one mapping, one VM
instance may be mapped onto one host machine and in many
to one mapping, many VM instances may be mapped onto
one host machine. Function f: Vi
Hi
describes theone to one
mapping and function f: Vi
Hi
maps manytoonemapping
from the host set HSa
based on the requirements and workload
of the use.In this way, VM instances may be mapped onto
host machine using the virtualization mapping diagram shown
in the figure [2].
VI. FUTURE DIRECTION
Cloud mapping model may provide good and efficient
quality services to the customers, in spite of that this model
has constraints and that could be taken as future work. As
hypervisor administrator has the control over the access of
VMs and it controls the monitoring system, socrashing down
the hypervisor or the failure of the hypervisor or attacks on
hypervisor administrator may lead to the performance
degradation. Dynamic mapping or run-time mapping
according to the requirements of the user make the whole
systemslow. So better mapping technique can be developed
for fast access.Each organization has their own application
interface, services and policies. Hence, in a cloud platform, it
is a mammoth task to maintain the heterogeneity of all
organization.
CONCLUSIONS
Rapid usage of Internet over the globe, Cloud Computing
has placed itself in everyfield of IT industry. To enhance our
mapping model, we mayfacilitate for the cloud users adding
or removing at run-time. Our model for virtualization
incorporated some new phases to improve its capabilities. It
may conclude that to facilitate maximum computation with
minimal physical resources our mappingmodel helps toutilize
the resources efficiently.
REFERENCES
[1] M. Armburst et al., “Above the Clouds: A Berkeley View of
Cloud Computing”, Tech. report, Univ. of California, Berkeley,
2009.
[2] RajkumarBuyyaa, Chee Shin Yeoa, , SrikumarVenugopala,
James Broberga, and IvonaBrandicc, “Cloud computing and
emerging IT platforms: Vision, hype, and reality for delivering
computing as the 5th
utility”, Future Generation Computer
Systems, Volume 25, Issue 6, June 2009, Pages 599-616.
[3] V. Sarathy, P. Narayan, RaoMikkilineni, “Next generation
cloud computing architecture -enabling real-time dynamism
for shared distributed physical infrastructure”, 19th IEEE
International Workshops on Enabling Technologies:
Infrastructures for Collaborative Enterprises (WETICE’10),
Larissa, Greece, 28-30 June 2010, pp. 48-53.
[4] RajkumarBuyya, Chee Shin Yeo, SrikumarVenugopal. Market-
Oriented Cloud Computing: Vision, Hype, and Reality for
Delivering IT Services as Computing Utilities. In 10th IEEE
15
7. Full Paper
©2012ACEEE
DOI:02.ACS.2012.3.
Proc. of Int. Conf. on Advances in Computer Science 2012
512
Models for the Next Generation of Software.
[11] “Cloud Computing Architectures” by HuagloryTianfield, 978-
1-4577-0653-0/11/$26.00 ©2011 IEEE.
[12] Souvik Pal and P.K.Pattnaik, “Efficient architectural
Framework of Cloud Computing”, “International Journal of
CloudComputing and Services Science (IJ-CLOSER)”, Vol.1,
No.2, June 2012, pp. 66~73.
[13] Judith Hurwitz, Robin Bloor, Marcia Kaufman, Fern Halper,
Cloud computing for dummies, Wiley Publications, Indiana,
2010.
[14] Jianfeng Yang et al, Cloud Computing Research and Security
Issues, 978-1-4244-5392-4/10, IEEE Proceedings, 2010.
[15] Souvik Pal, Suneeta Mohanty, SisirKunar Jena, Prasant Kumar
Pattnaik, “An Approach to Cross-Cloud Live Migration of
Virtual Machines in Cloud Computing Environment”,
Unpublished.
[16] B. Rochwerger et al, “The RESERVOIR model and architecture
for open federated cloud computing”, IBM Journal of Research
and Development, vol. 53, no. 4, 2009, pp. 1-11.
[17] PoojaMalgaonkar, RichaKoul, PriyankaThorat, MamtaZawar,
“Mapping of Virtual Machines in Private Cloud”, International
Journal of Computer Trends and Technology, volume2Issue2-
2011pp 54-57.
International Conference on High Performance Computing and
Communications, September 2008.pp 5–13.
[5 ] NaidilaSadashiv , S. M Dilip Kumar “Cluster, Grid and Cloud
Computing: A Detailed Comparison”, The 6th International
Conference on Computer Science & Education (ICCSE 2011)
August 3-5, 2011. SuperStar Virgo, Singapore, pages 477-482.
[6] K. Krauter, R. Buyya, and M. Maheswaran, “A Taxonomy
and Survey of Grid Resource Management Systems for
Distributed Computing”, Jr.of Software Practice and
Experience, 32,(2), pp. 135-164, 2002.
[7] M.Chetty and R. Buyya, “Weaving Computational Grids:
How Analogous Are They with Electrical Grids?”,Computing
in Science andEngineering (CiSE),4, pp. 61-71, 2002.
[8] Shuai Zhang et al “TheComparison Between Cloud Computing
and Grid Computing”, 2010 International Conference on
Computer Application and System Modeling (ICCASM 2010),
V11-72 to V11-75.
[9] L. Silva and R. Buyya, Parallel Programming Models and
Paradigms, High PerformanceCluster Computing:Programming
and Applications, RajkumarBuyya (editor), ISBN 0-13-
013785-5, Prentice Hall PTR, NJ, USA, 1999 .
[10] O’Reilly, T. What Is Web 2.0: Design Patterns and Business
16