Docstoc

CLOUD COMPUTING CHALLENGES AND RELATED SECURITY

Document Sample
CLOUD COMPUTING CHALLENGES AND RELATED SECURITY Powered By Docstoc
					CLOUD COMPUTING CHALLENGES AND RELATED
            SECURITY ISSUES
ABSTRACT:
The field of cloud computing is still in its
                                                    INTRODUCTION
infancy as far as implementation and usage,
partly because it is heavily promoted by
technology advancement and is so high resource      Cloud computing is not an innovation per se,
dependent that researches in academic               but a means to constructing IT services that
institutions have not had many opportunities to     use advanced computational power and
analyze and experiment with it. However,            improved storage capabilities. The main
CLOUD COMPUTING arises from the IT                  focus of cloud computing from the
technicians desire to add another layer of          provider's view as extraneous hardware
separation in processing information.               connected to support downtime on any
                                                    device in the network, without a change in
 At the moment, a general understanding of          the users' perspective. Also, the users'
cloud computing refers to the following             software image should be easily transferable
concepts: GRID COMPUTING, UTILITY
                                                    from one cloud to another. Balding proposes
COMPUTING, software as a service, storage in
the cloud and virtualization. These refer to a      that a layering mechanism should occur
client using a provider's service remotely, also    between the front-end software, middle-
known as in the cloud. Even if there is an          ware networking and back-end servers and
existent debate on whether those concepts           storage, so that each part can be designed,
should be separated and dealt with individually,    implemented, tested and ran independent
the general consensus is that all those terms       from subsequent layers. This paper
could be summarized by the cloud computing          introduces the current state of cloud
umbrella.                                           computing, with its development challenges,
                                                    academia and industry research efforts.
 Given its recent development and scarcity of       Further, it describes cloud computing
academic published work, many discussions on
                                                    security problems and benefits and
the topic of cloud security have surfaced from
engineers in companies that provide the             showcases a model of secure architecture for
aforementioned services. Nevertheless,              cloud computing implementation.
academia is developing in a significant presence,
being able to address numerous issues.
                                                    Cloud computing appeared in 2006, when
With the development of cloud computing, Data       Amazon’s Elastic Computing Cloud (EC2)
security becomes more and more important in         fires the world. Many information
cloud computing. This paper analyses the basic      Enterprises develop their platform for cloud
problem of cloud computing data security. With      computing. In 2007, Dell releases his
the analysis of HDFS architecture, we get the       solution of cloud computing, at the same
data security requirement of cloud computing        time IBM’s Blue Cloud comes in. Such as
and set up a mathematical data model for cloud
                                                    Google’s Mapreduce, Microsoftware’s
computing. Finally we build a data security
model for cloud computing.                          Windows Azure .According to an estimation
                                                    , by 2012, the Cloud computing market
Key Terms- Cloud Computing, Grid                    should reach $420 billion. All this have
Computing, Utility Computing, SaaS, HDFs and        show the coming of the epoch time of cloud
Data Security                                       computing.
model.
The emergence of the Cloud system has             security measures on the edge of the
simplified the deployment of large-scale          hardware platform, while cloud computing
distributed systems for software vendors.         may be a server in a number of virtual
The Cloud system provides a simple and            servers, the virtual server may belong to
unified interface between vendor and user,        different logical server group, virtual server,
allowing vendors to focus more on the             therefore there is the possibility of attacking
software itself rather than the underlying        each other ,which brings virtual servers a lot
framework. Applications on the Cloud              of security threats. Virtual machine
include Software as a Service system and          extending the edge of clouds makes the
Multi-tenant databases . The Cloud system         disappearance of the network boundary,
dynamically allocates computational               thereby affecting almost all aspects of
resources in response to customers’ resource      security, the traditional physical isolation
reservation requests and in accordance with       and hardware-based security infrastructure
customers’ predesigned quality of service.        cannot stop the clouds computer
                                                  environment of mutual attacks between the
Risk coming with opportunity, the problem         virtual machine.
of data security in Cloud computing become
bottleneck of cloud computing. In this paper
we want to set up a security model for cloud
computing , the rest of the paper is              The Existence of Super-user:
organized as follows: We present the data         For the enterprise providing cloud
security problem of cloud computing in the        computing services, they have the right to
next section and then discuss the details of      carry out the management and maintenance
requirement of security.                          of data, the existence of super-users to
                                                  greatly simplify the data management
Windows Azure, the virtual machine                function, but it is a serious threat to user
technology is considered as a cloud               privacy. Super-powers is a double edged
computing platform of the fundamental             sword, it brings convenience to users and at
component, the differences between Blue           the same time poses a threat to users. In an
Cloud and Windows Azure is that virtual           era of personal privacy, personal data should
machine running on Linux operating system         be really protected, and the fact that cloud
or Microsoft Windows operating system.            computing platform to provide personal
Virtual Machine technology bring obvious          services in the confidentiality of personal
advantages, it allows the operation of the        privacy on the existence of defects. Not only
server which is no longer dependent on the        individual users but also the organizations
physical device, but on the virtual servers. In   have similar potential threats, e.g.
virtual machine, a physical change or             corporate users and trade secrets stored in
imigration does not affect the services           the cloud computing platform may be stolen.
provided by the service provider. if user         therefore the use of super user rights must be
need more services, the provider can meet         controlled in the cloud.
user’s needs without having to concern the
physical hardware.
                                                  Consistency of Data:
However, the virtual server from the logical      Cloud environment is a dynamic
server group brings a lot of security             environment, where the user's data transmits
problems. The traditional data center             from the data centre to the user's client. For
the system, the user's data is changing all the   Google are currently providing such
time. Read and write data relating to the         services, charging clients using an on-
identity of the user authentication and           demand policy. references statistics that
permission issues. In a virtual machine,          suggest one third of breaches are due to
there may be different users’ data which          laptops falling in the wrong hands and about
must be strict managed.                           16% due to stolen items by employees.
                                                  Storing the data in the cloud can prevent
                                                  these issues altogether. Moreover, vendors
The traditional model of access control is
                                                  can update application/OS/middleware
built in the edge of computers, so it is weak
                                                  security patches faster because of higher
to control reading and writing among
                                                  availability of staff and resources.
distributed computers. It is clear that
traditional access control is obviously not
                                                  According to cloud vendors, most thefts
suitable for cloud computing environments.
                                                  occur when users with authorized access do
In the cloud computing environment, the
                                                  not handle data appropriately. Upon a logout
traditional access control mechanism has
                                                  from the cloud session, the browser may be
serious shortcomings.
                                                  configured to delete data automatically and
New Technology:                                   log files on the vendor side indicate which
                                                  user accessed what data. This approach may
The concept of cloud computing is built on        be deemed safer that storing data on the
new architecture. The new architecture            client side. There are some applications for
comprised of a variety of new technologies,       which cloud computing is the best option.
such as Hadoop, Hbase, which enhances the         One example is the New York Times using
performance of cloud systems but brings in        Amazon's cloud service to generate PDF
risks at the same time. In the cloud              documents of several-decade old articles.
environment, users create many dynamic
virtual organizations, first set up in co-        The estimated time for doing the task on the
operation usually occurs in a relationship of     Times' servers was 14 years, whereas the
trust between organizations rather than           cloud provided the answer in one day for a
individual level. So those users based on the     couple hundred dollars. However, the profile
expression of restrictions on the basis of        of the companies that currently use the cloud
proof strategy is often difficult to follow;      technology includes Web 2.0 start-ups that
which frequently occurs in many of the            want to minimize material cost, application
interactive nodes between the virtual             developers that want to enable their software
machine, and is dynamic, unpredictable.           as a service or enterprises that are exploring
Cloud computing environment provides a            the cloud with trivial applications. The fact
user of the "buy" the full access to resources    that cloud computing is not used for all of its
which has also increased security risks.          potential is due to a variety of concerns. The
                                                  following surveys the market in terms of
Critics argue that cloud computing is not         continuous innovation, academia and
secure enough because data leaves                 industry research efforts and cloud
companies' local area networks. It is up to       computing challenges.
the clients to decide the vendors, depending
on how willing they are to implement secure
policies and be subject to 3rd party
verifications. Sales force, Amazon and            REQUIREMENT OF SECURITY
Following, with the analysis of the widely        system still support the creation, delete,
used cloud computing technology--HDFS             open, close, read, write and other operations
(Hadoop Distributed File System), we will         on files.
get the data security needs of cloud
computing. HDFS is used in large-scale            By analyzing of HDFS, data security needs
cloud computing in a typical distributed file     of cloud computing can be divided into the
system architecture, its design goal is to run    following points:
on commercial hardware, due to the support
of Google, and the advantages of open                   The client authentication
source, it has been applied in the basis of              requirements in login: The vast
cloud facilities. HDFS is very similar to the            majority of cloud computing through
existing distributed file system, such as GFS            a browser client, such as IE, and the
(Google File System); they have the same                 user’s identity as a cloud computing
objectives, performance, availability and                applications demand for the primary
stability. HDFS initially used in the Apache             needs.
Nutch web search engine and become the                  The existence of a single point of
core of Apache Hadoop project.                           failure in Name node: if name node
                                                         is attacked or failure, there will be
                                                         disastrous consequences on the
                                                         system. So the effectiveness of Name
                                                         node in cloud computing and its
                                                         efficiency is key to the success of
                                                         data protection, so to enhance Name
                                                         node’s security is very important.

                                                        The rapid recovery of data blocks
                                                         and r/w rights control: Data node is a
                                                         data storage node, there is the
                                                         possibility of failure and cannot
                                                         guarantee the availability of data.
                                                         Currently each data storage block in
                                                         HDFS has at least 3 replicas, which
                                                         is HDFS’s backup strategy. When
                                                         comes to how to ensure the safety of
                                                         reading and writing data, HDFS has
                                                         not made any detailed Explanation,
HDFS used the master/slave backup mode.                  so the needs to ensure rapid recovery
As shown in Figure 1. The master is called               and to make reading and writing data
Name node, which manages the file system                 operation fully controllable cannot
name space and controls access to the client.            be ignored.
Other slave nodes is called Data node, Data
node controls access to his client. In this             In addition to the above three
storage system, a file is cut into small pieces          requirements, the other, such as
of paper, Name node maps the file blocks to              access control, file encryption, such
Data nodes above. While HDFS does not                    as demand for cloud computing
have the POSIX compatibility, the file
       model for data security issues must        Data Security Model:
       be taken into account.
                                                  Data model of cloud computing can be
                                                  described in math as follows:
DATA SECURITY MODEL
                                                        The model used three-level defense
Principle of Data Security:                              system structure, in which each floor
                                                         performs its own duty to ensure that
All the data security technique is built on              the data security of cloud layers.
confidentiality, integrity and availability of         The first layer: responsible for user
these three                                              authentication, the user of digital
basic principles. Confidentiality refers to the          certificates issued by the appropriate,
so-called hidden the actual data or                      manage user permissions;
information, especially in the military and            The second layer: responsible for
other sensitive areas, the confidentiality of            user's data encryption, and protect
data on the more stringent requirements. For             the privacy of users through a certain
cloud computing, the data are stored in "data            way;
center", the security and confidentiality of           The third layer: The user data for fast
user data is even more important. The so-                recovery, system protection is the
called integrity of data in any state is                 last layer of user data.
not subject to the need to guarantee                   With three-level structure, user
unauthorized deletion, modification or                   authentication is used to ensure that
damage. The availability of data means that              data is not tampered. The user
users can have the expectations of the use of            authenticated can manage the data by
data by the use of capacity.                             operations: Add, modify, delete and
                                                         so on. If the user authentication
                                                         system is deceived by illegal means,
                                                         and malign user enters the system,
                                                         file encryption and privacy
                                                         protection can provide this level of
                                                         defense. In this layer user data is
                                                         encrypted, even if the key was the
                                                         illegally accessed, through privacy
                                                         protection, malign user will still be
                                                         not unable to obtain effective access
                                                         to information, which is very
                                                         important to protect business users’
                                                         trade secrets in cloud computing
                                                         environment. Finally, the rapid
                                                         restoration of files layer, through fast
                                                         recovery algorithm, makes user data
                                                         be able to get the maximum recovery
                                                         even in case of damage.
                                                  Security refers to confidentiality, integrity
                                                  and availability, which pose major issues for
                                                  cloud vendors. Confidentiality refers to who
stores the encryption keys - data from            point. Also, authentication should be backed
company A, stored in an encrypted format at       by several methods like password plus flash
company B must be kept secure from                card, or password plus finger print, or some
employees of B; thus, the client company          combination of external hardware and
should own the encryption keys. Integrity         password. One benefit of cloud computing is
refers to the face that no common policies        that client software security does not need to
exist for approved data exchanges; the            be enforced as strictly as before. This aspect
industry has various protocols used to push       concerns the view of cloud computing as
different software images or jobs. One way        software as a service, as it becomes more
to maintain data security on the client side is   important to ensure security of data transfer
the use of thin clients that run with as few      rather than a traditional secure application
resources as possible and do not store any        life cycle.
user data, so passwords cannot be stolen.
The concept seems to be impervious to
attacks based on capturing this data.
However, companies have implemented               Cloud Computing Challenges:
systems with unpublished APIs, claiming
that it improves security; unfortunately, this           Challenges that cloud computing
can be reversed engineered; also, using                  currently faces in being deployed on
DHCP and FTP to perform tasks such as                    a large enterprise scale:
firmware upgrades has long been rendered
as insecure. Nevertheless, products from                        Self-healing - in case of
Wyse are marketed with their thin client as                      application/network/data
one of the safest, by using those exact                          storage failure, there will
features.                                                        always be a backup running
                                                                 without major delays, making
                                                                 the resource switch appear
                                                                 seamless to the user.
                                                                SLA-driven - cloud is
                                                                 administrated by service level
                                                                 agreements that allow several
                                                                 instances of one application
                                                                 to be replicated on multiple
                                                                 servers if need arises;
                                                                 dependent on a priority
                                                                 scheme, the cloud may
                                                                 minimize or shut down a
                                                                 lower level application.
Lastly, the most problematic issue is
                                                                Multi-tenancy - the cloud
availability, as several companies using
                                                                 permits multiple clients to
cloud computing have already experienced
                                                                 use the same hardware at the
downtime (Amazon servers subject to what
                                                                 same time, without them
appeared to be a denial of service attack).
                                                                 knowing it, possibly causing
Other things to keep in mind are contract
                                                                 conflicts of interest among
policies between clients and vendors, so that
                                                                 customers.
data belongs only to the client at all times,
preventing third parties to be involved at any
              Service-oriented - cloud         encouraged to pursue cloud computing as
               allows one client to use         the alternative to supporting their own
               multiple applications in         hardware backbone.
               creating its own.
              Virtualized - applications are
               not hardware specific;
               various programs may run on
               one machine using
               virtualization or many
               machines may run one
               program.
              Linearly scalable - cloud
               should handle an increase in
               data processing linearly; if
               "n" times more users need a
               resource, the time to
               complete the request with "n"     However cloud computing does not come
               more resources should be         without its pitfalls. For starters, a cloud is a
               roughly the same.                single point of failure for multiple resources.
              Data management -                Even though network carriers such as AT&T
               distribution, partitioning,      believe a distributed cloud structure is the
               security and synchronization     right implementation, it faces major
               of data.                         challenges in finding the optimal approach
                                                for low power transmission and high
                                                network availability [Croll08]; some people
                                                believe that major corporations will shy
                                                away from implementing cloud solutions in
                                                the near future due to ineffective security
                                                policies. One problem comes from the fact
                                                that different cloud providers have different
                                                ways to store data, so creating a distributed
                                                cloud implies more challenges to be solved
                                                between vendors.



                                                Cloud Computing Security Issues:
Security Challenges:
                                                Identified seven issues that need to be
Start-up companies often lack the protection    addressed before enterprises consider
measures to weather off an attack on their      switching to the cloud computing model.
servers due to the scarcity of resources -      They are as follows:
poor programming that explores software
vulnerabilities (PHP, JavaScript, etc) open            Privileged user access - information
ports to firewalls or inexistent load-balance           transmitted from the client through
algorithms susceptible to denial of service             the Internet poses a certain degree of
attacks. For this reason, new companies are
    risk, because of issues of data                   data if the current provider is bought
    ownership; enterprises should spend               out by another firm Given that not all
    time getting to know their providers              of the above need to be improved
    and their regulations as much as                  depending on the application at hand,
    possible before assigning some                    it is still paramount that consensus is
    trivial applications first to test the            reached on the issues regarding
    water.                                            standardization.




    Fig:Security in cloud
   Regulatory compliance - clients are
    accountable for the security of their
    solution, as they can choose between
    providers that allow to be audited by
    3rd party organizations that check         Security Benefits:
    levels of security and providers that
    don't .                                    There are definitely plenty of concerns
                                               regarding the inability to trust cloud
   Data location - depending on
                                               computing due to its security issues.
    contracts, some clients might never
                                               However, cloud computing comes with
    know what country or what
                                               several benefits that address data security.
    jurisdiction their data is located .
                                               The following sections looks into addressing
   Data segregation - encrypted
                                               concepts such as centralized data, incident
    information from multiple
                                               response or logging.
    companies may be stored on the
    same hard disk, so a mechanism to
                                               Centralized Data refers to the approach of
    separate data should be deployed by
                                               placing all eggs in one basket. It might be
    the provider.
                                               dangerous to think that if the cloud goes
         recovery - every provider
                                               down, so does the service they provide, but
             should have a disaster
                                               at the same time, it is easier to monitor.
             recovery protocol to protect
                                               Storing data in the cloud voids many issues
             user data .
                                               related to losing laptops or flash drives,
         investigative support - if a
                                               which has been the most common way of
             client suspects faulty activity
                                               loosing data for large enterprises or
             from the provider, it may not
                                               government organizations. The laptop would
             have many legal ways pursue
                                               only store a small cache to interface with the
             an investigation
                                               thin client, but the authentication is done
   Long-term viability - refers to the        through the network, in the cloud. In
    ability to retract a contract and all      addition to this, when a laptop is known to
be stolen, administrators can block its          worry about storage space for log files and
attempted access based on its identifier or      enjoys a faster way of searching through
MAC address. Moreover, it is easier and          them. Moreover, it allows for a convenient
cheaper to store data encrypted in the cloud     way to observe which user accessed certain
that to perform disk encryption on every         resources at any given time.
piece of hardware or backup tape.
                                                 Improvement of Secure Software refers to
                                                 several aspects in the development lifecycle
                                                 of a product. Initially, a company that is
                                                 thinking of placing their application in the
                                                 cloud knows that the cost of running the
                                                 application are directly proportional with the
                                                 number of processing cycles, thus creating
                                                 an incentive for an optimal implementation.
                                                 Secondly, it becomes easier to monitor the
                                                 effects of various security policies
                                                 implemented in the software, without the
                                                 overhead of traditional switching
                                                 environments from development to
                                                 production or to testing. Creating a new
Incident Response refers to the ability to       environment simply means creating a clone
procure a resource such as a database server     of the extant one. Thirdly, software run
or supercomputing power or use a testing         behind an architecture that is build for
environment whenever needed. This                secure transactions at a physical, data link,
bypasses the supplemental red tape               network and transport layer, making it easier
associated with traditional requesting of        to design the application without the
resources within the corporate world. Also,      outspoken need of a security software
if a server is down for re-imaging or disk       engineer. Moreover, some cloud providers
clean-up, the client may easily create similar   may use code scanning to detect
instances of their environment on other          vulnerabilities in the application code.
machines, improving the acquisition time.
From a security standpoint, cloud providers
already provide algorithms for generating
hashes or checksums whenever a file is           Secure Architecture Models:
stored in the cloud, which bypasses the
local/client need for encrypting. This does      Open Security Architecture (OSA) provides
not imply that clients should not encrypt the    free frameworks that are easily integrated in
data before sending it, but merely that the      applications, for the security architecture
service is already in place for them.            community. Its patterns are based on
                                                 schematics that show the information traffic
Password Assurance Testing is a service that     flow for a particular implementation as well
can be used to harness the computational         as policies implemented at each step for
power of the cloud in attempts to break into     security reasons. The following description
a company's system by guessing passwords.        of a proposed cloud computing architecture,
This approach minimizes resources and time       also shown in fig. 2, should help the reader
spent on the client side. Logging benefits       envision the components of cloud
come from the idea that the client need not
computing architectures along with               itself. The following summary looks at their
descriptions of elements that make it secure.    attributions and mechanisms available for
The important entities involved in the data      them.
flow are end users, developers, system
architect, 3rd party auditors and the cloud




            Figure 2. Cloud Computing Model - Open Secure Architecture
End Users:                                       sessions are not being altered by a man in
                                                 the middle attack. At a lower level, the
End Users need to access certain resources       network can be made secure by the use of
in the cloud and should be aware of access       secure internet protocol (IPsec). Lastly, the
agreements such as acceptable use or             final middle point between end users and the
conflict of interest. In this model, end user    cloud is transmission confidentiality or the
signatures may be used to confirm someone        guarantee that no one is listening on the
is committed to such policies. The client        conversation between authenticated users
organization should run mechanisms to            and the cloud. The same mechanisms
detect vulnerable code or protocols at entry     mentioned above can also guarantee
points such as firewalls, servers, or mobile     confidentiality.
devices and upload patches on the local
systems as soon as they are found. Thus, this    System Architects:
approach ensures security on the end users
and on the cloud alike. However, the cloud       System architects are employed with writing
needs to be secure from any user with            the policies that pertain to the installation
malicious intent that may attempt to gain        and configuration of hardware components
access to information or shut down a             such as firewalls, servers, routers, and
service.                                         software such as operating systems, thin
                                                 clients, etc. They designate control protocols
                                                 to direct the information flow within the
                                                 cloud such as router update/queuing
                                                 protocols, proxy server configurations or
                                                 encrypted tunnels.

                                                 Developers:

                                                 Developers building an application in the
                                                 cloud need to access the infrastructure where
                                                 the development environment is located.
                                                 They also need to access some configuration
                                                 server that allows them to test applications
For this reason, the cloud should include a      from various views. Cloud computing can
denial of service (DOS) protection. One way      improve software development by scaling
of enforcing DOS protection is done by           the software environment through elasticity
improving the infrastructure with more           of resources. For example, one developer
bandwidth and better computational power         can get extra hard space as an on-demand
which the cloud has abundantly. However,         resource, instead of placing a work order
in the more traditional sense, it involves       and wait for several days for the permission.
filtering certain packets that have similar IP   Developers may desire extra virtual
source addresses or server requests. The         machines to either generate test data or to
next issue concerning the cloud provider to      perform data analysis, processes which take
end users is transmission integrity.             significant time.

 One way of implementing integrity is by
using secure socket layer (SSL) or transport
layer security (TLS) to ensure that the
                                                 implementation for a large-sized enterprise.
                                                 IT technicians are spearheading the
                                                 challenge, while academia is bit slower to
                                                 react. Several groups have recently been
                                                 formed, such as the Cloud Security Alliance
                                                 or the Open Cloud Consortium, with the
                                                 goal of exploring the possibilities offered by
                                                 cloud computing and to establish a common
                                                 language among different providers. In this
                                                 boiling pot, cloud computing is facing
                                                 several issues in gaining recognition for its
                                                 merits. Its security deficiencies and benefits
                                                 need to be carefully weighed before making
                                                 a decision to implement it. However, the
 Also, using more processing power from          future looks less cloudy as far as more
the cloud can help in catching up with the       people being attracted by the topic and
development schedule. The cloud also helps       pursuing research to improve on its
developers create multiple evaluation            drawbacks.
versions environments for their applications,
bypassing the need to incorporate additional     As the development of cloud computing,
security within the application and placing      security issue has become a top priority.
the burden on the cloud provider. One            This paper discusses the cloud computing
significant drawback of cloud computing at       environment with the safety issues through
the moment is its limitations to Intel x86       analyzing a cloud computing framework
processor architecture. Even if this may very    HDFS’s security needs. Finally we conclude
well change in the future, it is another         a cloud computing model for data security.
stumbling block that developers and cloud
computing experts need to overcome.
Software monitoring may be done by               References:
monitoring API calls for server requests.
With an architectural model where data is        [1] Rajkumar Buyya Market-Oriented Cloud
centralized, all eyes are focused in one         Computing:Vision,Hype,and Reality for
direction, which implies better monitoring,      Delivering IT Services as Computing
although ultimately the issue rests with the     Utilities 2008
developers/clients on how much effort will       [2] Jean-Daniel Cryans, Criteria to Compare
be directed in this regard. As far as security   Cloud Computing with Current Database
patches for the software as service approach,    technology 2008
updating a patch is easier done in the cloud     [3] Christopher Moretti,All-Pairs: An
and shared with everyone seamlessly, rather      Abstraction for Data-Intensive Cloud
than finding every machine that has the          Computing IEEE 2008
software installed locally.                      [4] Huan Liu, Dan Orban��GridBatch:
                                                 Cloud Computing for Large-Scale Data-
Conclusion :                                     Intensive Batch Applications IEEE DOI
                                                 10.1109/CCGRID.2008.30
Cloud computing is still struggling in its       [5] Mladen A. Vouk��Cloud Computing –
infancy, with positive and negative              Issues, Research and Implementations
comments made on its possible
Journal of Computing and Information    [8] Greg Boss, Cloud Computing IBM
Technology - CIT 16, 2008, 4, 235–246   2007.10
[6] Bob Gourley ��Cloud Computing and    [9] Jeffrey Dean and Sanjay Ghemawat
Net Centric Operations ��Department of   MapReduce: Simplied Data Processing on
Defense Information Management and      Large Clusters Google, Inc 2004.
Information Technology Strategic Plan   [10] David Chappell ,Introducing the Azure
2008-2009                               Services Platform October 2008
[7] Cloud Computing Security:making
Virtual Machines Cloud-Ready

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:8
posted:12/2/2011
language:English
pages:14