Market-Oriented Cloud Computing:

Vision, Hype, and Reality for Delivering IT Services as Computing Utilities

S.S.P.M's College Of Engineering

Department Of Computer Engineering
The University Of Mumbai

Ritesh Lad Shailesh Dhavan

email id:[email protected] email id:[email protected]

Abstract sophisticated, we will probably see the spread of

This keynote paper: presents a 21st century vision ‘computer utilities’ which, like present electric and
of computing; identifies various computing paradigms telephone utilities, will service individual homes and
promising to deliver the vision of computing utilities; offices across the country.” This vision of the
defines Cloud computing and provides the architecture computing utility based on the service provisioning
for creating market-oriented Clouds by leveraging model anticipates the massive transformation of the
technologies such as VMs; provides thoughts on entire computing industry in the 21st century whereby
market-based resource management strategies that computing services will be readily available on
encompass both customer-driven service management demand, like other utility services available in today’s
and computational risk management to sustain SLA- society. Similarly, computing service users
oriented resource allocation; presents some (consumers) need to pay providers only when they
representative Cloud platforms especially those access computing services. In addition, consumers no
developed in industries along with our current work longer need to invest heavily or encounter difficulties
towards realising market-oriented resource allocation in building and maintaining complex IT infrastructure.
of Clouds by leveraging the 3rd generation Aneka Software practitioners are facing numerous new
enterprise Grid technology; reveals our early thoughts challenges toward creating software for millions of
on interconnecting Clouds for dynamically creating an consumers to use as a service rather than to run on their
atmospheric computing environment along with individual computers. Over the years, new computing
pointers to future community research; and concludes paradigms have been proposed and adopted, with the
with the need for convergence of competing IT emergence of technological advances such as multi-
paradigms for delivering our 21st century vision. core processors and networked computing
environments, to edge closer toward achieving this
1. Introduction grand vision. As shown in Figure 1, these new
computing paradigms include cluster computing, Grid
With the advancement of the modern human society,
computing, P2P computing, service computing,
basic essential services are commonly provided such
market-oriented computing, and most recently Cloud
that everyone can easily obtain access to them. Today,
computing. All these paradigms promise to provide
utility services, such as water, electricity, gas, and
certain attributes or capabilities in order to realize the
telephony are deemed necessary for fulfilling daily life
possibly 1 trillion dollars worth of the utility/pervasive
routines. These utility services are accessed so
computing industry as quoted by Sun Microsystems
frequently that they need to be available whenever the
co-founder Bill Joy [2]. Computing services need to be
consumer requires them at any time. Consumers are
highly reliable, scalable, and autonomic to support
then able to pay service providers based on their usage
ubiquitous access, dynamic discovery and
of these utility services.
composability. In particular, consumers can determine
In 1969, Leonard Kleinrock [1], one of the chief the required service level through Quality of Service
scientists of the original Advanced Research Projects (QoS) parameters and Service Level Agreements
Agency Network (ARPANET) project which seeded (SLAs). Of all these computing paradigms, the two
the Internet, said: “As of now, computer networks are most promising ones appear to be Grid computing and
still in their infancy, but as they grow up and become Cloud computing.
 } ?
Web
Data Centres
Utility Computing
Service Computing
Grid Computing +
P2P Computing
Cloud Computing
Market-Oriented
•Ubiquitous
Computing access
-Trillion $ business
… •Reliability - Who will own it?
•Scalability
Paradigms •Autonomic
•Dynamic
discovery
•Composability
•QoS
•SLA
•…
Attributes/Capabilities

Figure 1: Various paradigms promising to deliver IT as services.

A Grid [3] enables the sharing, selection, and Cloud infrastructure is very robust and will always be
aggregation of a wide variety of geographically available at any time.
distributed resources including supercomputers,
storage systems, data sources, and specialized devices 1.1 Definition and Trends
owned by different organizations for solving large-
A number of computing researchers and practitioners
scale resource-intensive problems in science,
have attempted to define Clouds in various ways [6].
engineering, and commerce. Inspired by the electrical
Based on our observation of the essence of what
power Grid’s pervasiveness, ease of use, and reliability
Clouds are promising to be, we propose the following
[4], the motivation of Grid computing was initially
definition:
driven by large-scale, resource (computational and
data)-intensive scientific applications that required • "A Cloud is a type of parallel and distributed
more resources than a single computer (PC, system consisting of a collection of inter-
workstation, supercomputer, or cluster) could have connected and virtualised computers that are
provided in a single administrative domain. Due to its dynamically provisioned and presented as one
potential to make impact on the 21st century as much or more unified computing resources based on
as the electric power Grid did on the 20th century, Grid service-level agreements established through
computing has been hailed as the next revolution after negotiation between the service provider and
the Internet and the Web. consumers.”
Today, the latest paradigm to emerge is that of At a cursory glance, Clouds appear to be a
Cloud computing [5] which promises reliable services combination of clusters and Grids. However, this is not
delivered through next-generation data centers that are the case. Clouds are clearly next-generation data
built on compute and storage virtualization centers with nodes “virtualized” through hypervisor
technologies. Consumers will be able to access technologies such as VMs, dynamically “provisioned”
applications and data from a “Cloud” anywhere in the on demand as a personalized resource collection to
world on demand. In other words, the Cloud appears to meet a specific service-level agreement, which is
be a single point of access for all the computing needs established through a “negotiation” and accessible as a
of consumers. The consumers are assured that the composable service via “Web 2.0” technologies.
 grid

cloud

cluster
cloud

Legend: Cluster computing, Grid computing, Cloud computing

Figure 2: Google search trends for the last 12 months.

1.2 Web Search Trends 2. Market-Oriented Cloud Architecture

The popularity of different paradigms varies with time.
As consumers rely on Cloud providers to supply all
The Web search popularity, as measured by the Google
their computing needs, they will require specific QoS
search trends during the last 12 months, for terms
to be maintained by their providers in order to meet
“cluster computing”, “Grid computing”, and “Cloud
their objectives and sustain their operations. Cloud
computing” is shown in Figure 2. From the Google
providers will need to consider and meet different QoS
trends, it can be observed that cluster computing was a
parameters of each individual consumer as negotiated
popular term during 1990s, from early 2000 Grid
in specific SLAs. To achieve this, Cloud providers can
computing become popular, and recently Cloud
no longer continue to deploy traditional system-centric
computing started gaining popularity.
resource management architecture that do not provide
Spot points in Figure 2 indicate the release of news incentives for them to share their resources and still
related to Cloud computing as follows: regard all service requests to be of equal importance.
IBM Introduces 'Blue Cloud' Computing, CIO Instead, market-oriented resource management [7] is
Today - Nov 15 2007 necessary to regulate the supply and demand of Cloud
IBM, EU Launch RESERVOIR Research Initiative resources at market equilibrium, provide feedback in
for Cloud Computing, IT News Online - Feb 7 terms of economic incentives for both Cloud
2008 consumers and providers, and promote QoS-based
resource allocation mechanisms that differentiate
Google and Salesforce.com in Cloud computing service requests based on their utility.
deal, Siliconrepublic.com - Apr 14 2008
Figure 3 shows the high-level architecture for
Demystifying Cloud Computing, Intelligent supporting market-oriented resource allocation in Data
Enterprise - Jun 11 2008 Centers and Clouds. There are basically four main
Yahoo realigns to support Cloud computing, 'core entities involved:
strategies', San Antonio Business Journal - Jun 27 • Users/Brokers: Users or brokers acting on
2008 their behalf submit service requests from
Merrill Lynch Estimates "Cloud Computing" To anywhere in the world to the Data Center and
Be $100 Billion Market, SYS-CON Media - Jul 8 Cloud to be processed.
2008 • SLA Resource Allocator: The SLA Resource
Allocator acts as the interface between the
Data Center/Cloud service provider and
 Figure 3: High-level market-oriented cloud architecture.

external users/brokers. It requires the effectively. Then, it assigns requests

interaction of the following mechanisms to to VMs and determines resource
support SLA-oriented resource management: entitlements for allocated VMs.
o Service Request Examiner and o Pricing: The Pricing mechanism
Admission Control: When a service decides how service requests are
request is first submitted, the Service charged. For instance, requests can
Request Examiner and Admission be charged based on submission time
Control mechanism interprets the (peak/off-peak), pricing rates
submitted request for QoS (fixed/changing) or availability of
requirements before determining resources (supply/demand). Pricing
whether to accept or reject the serves as a basis for managing the
request. Thus, it ensures that there is supply and demand of computing
no overloading of resources whereby resources within the Data Center and
many service requests cannot be facilitates in prioritizing resource
fulfilled successfully due to limited allocations effectively.
resources available. It also needs the
o Accounting: The Accounting
latest status information regarding
mechanism maintains the actual
resource availability (from VM
usage of resources by requests so
Monitor mechanism) and workload
that the final cost can be computed
processing (from Service Request
and charged to the users. In addition,
Monitor mechanism) in order to
the maintained historical usage
make resource allocation decisions
information can be utilized by the
 Service Request Examiner and • support customer-driven service management
Admission Control mechanism to based on customer profiles and requested
improve resource allocation service requirements,
decisions.
• define computational risk management tactics
o VM Monitor: The VM Monitor to identify, assess, and manage risks involved
mechanism keeps track of the in the execution of applications with regards
availability of VMs and their to service requirements and customer needs,
resource entitlements.
• derive appropriate market-based resource
o Dispatcher: The Dispatcher management strategies that encompass both
mechanism starts the execution of customer-driven service management and
accepted service requests on computational risk management to sustain
allocated VMs. SLA-oriented resource allocation,
o Service Request Monitor: The • incorporate autonomic resource management
Service Request Monitor mechanism models that effectively self-manage changes
keeps track of the execution progress in service requirements to satisfy both new
of service requests. service demands and existing service
• VMs: Multiple VMs can be started and obligations, and
stopped dynamically on a single physical • leverage VM technology to dynamically
machine to meet accepted service requests, assign resource shares according to service
hence providing maximum flexibility to requirements.
configure various partitions of resources on
the same physical machine to different
specific requirements of service requests. In 3. Emerging Cloud Platforms
addition, multiple VMs can concurrently run Industry analysts have made bullish projections on how
applications based on different operating Cloud computing will transform the entire computing
system environments on a single physical industry. According to a recent Merrill Lynch research
machine since every VM is completely note [9], Cloud computing is expected to be a “$160-
isolated from one another on the same billion addressable market opportunity, including $95-
physical machine. billion in business and productivity applications, and
another $65-billion in online advertising”. Another
• Physical Machines: The Data Center
research study by Morgan Stanley [10] has also
comprises multiple computing servers that
identified Cloud computing as one of the prominent
provide resources to meet service demands.
technology trends. As the computing industry shifts
In the case of a Cloud as a commercial offering to toward providing Platform as a Service (PaaS) and
enable crucial business operations of companies, there Software as a Service (SaaS) for consumers and
are critical QoS parameters to consider in a service enterprises to access on demand regardless of time and
request, such as time, cost, reliability and trust/security. location, there will be an increase in the number of
In particular, QoS requirements cannot be static and Cloud platforms available. Recently, several academic
need to be dynamically updated over time due to and industrial organisations have started investigating
continuing changes in business operations and and developing technologies and infrastructure for
operating environments. In short, there should be Cloud Computing. Academic efforts include Virtual
greater importance on customers since they pay for Workspaces [11] and OpenNebula [12]. In this section,
accessing services in Clouds. In addition, the state-of- we compare six representative Cloud platforms with
the-art in Cloud computing has no or limited support industrial linkages in Table 1.
for dynamic negotiation of SLAs between participants
and mechanisms for automatic allocation of resources Amazon Elastic Compute Cloud (EC2) [13]
to multiple competing requests. Recently, we have provides a virtual computing environment that enables
developed negotiation mechanisms based on alternate a user to run Linux-based applications. The user can
offers protocol for establishing SLAs [8]. These have either create a new Amazon Machine Image (AMI)
high potential for their adoption in Cloud computing containing the applications, libraries, data and
systems built using VMs. associated configuration settings, or select from a
library of globally available AMIs. The user then needs
Commercial offerings of market-oriented Clouds
to upload the created or selected AMIs to Amazon
must be able to:
 Table 1: Comparison of some representative Cloud platforms.
System
Amazon Elastic Sun
Google Microsoft GRIDS Lab
Compute Cloud Network.com
App Engine Live Mesh Aneka
(EC2) (Sun Grid)
Property
Software
Platform for
Focus Infrastructure Platform Infrastructure Infrastructure
enterprise
Clouds
Compute, Storage Web
Service Type Storage Compute Compute
(Amazon S3) application
Job
OS Level running Resource
Application management
Virtualisation on a Xen OS level Manager and
container system (Sun
hypervisor Scheduler
Grid Engine)
Dynamic SLA-based
Negotiation of Resource
None None None None
QoS Reservation on
Parameters Aneka side.
Web-based
Job submission
Amazon EC2 Web-based Live Desktop Workbench,
User Access scripts, Sun
Command-line Administration and any devices Web-based
Interface Grid Web
Tools Console with Live Mesh portal
portal
installed
Web APIs Yes Yes Unknown Yes Yes

Value-added
Service Yes No No Yes No
Providers
APIs supporting
different
Customizable
Solaris OS, programming
Programming Linux-based
Python Not applicable Java, C, C++, models in C#
Framework Amazon Machine
FORTRAN and other .Net
Image (AMI)
supported
languages
Simple Storage Service (S3), before he can start, stop, data that can be accessed across required devices (such
and monitor instances of the uploaded AMIs. Amazon as computers and mobile phones) from anywhere in the
EC2 charges the user for the time when the instance is world. The user is able to access the uploaded
alive, while Amazon S3 charges for any data transfer applications and data through a Web-based Live
(both upload and download). Dekstop or his own devices with Live Mesh software
Google App Engine [14] allows a user to run Web installed. Each user’s Live Mesh is password-protected
applications written using the Python programming and authenticated via his Windows Live Login, while
language. Other than supporting the Python standard all file transfers are protected using Secure Socket
library, Google App Engine also supports Application Layers (SSL).
Programming Interfaces (APIs) for the datastore, Sun network.com (Sun Grid) [16] enables the user
Google Accounts, URL fetch, image manipulation, and to run Solaris OS, Java, C, C++, and FORTRAN based
email services. Google App Engine also provides a applications. First, the user has to build and debug his
Web-based Administration Console for the user to applications and runtime scripts in a local development
easily manage his running Web applications. environment that is configured to be similar to that on
Currently, Google App Engine is free to use with up to the Sun Grid. Then, he needs to create a bundled zip
500MB of storage and about 5 million page views per archive (containing all the related scripts, libraries,
month. executable binaries and input data) and upload it to Sun
Microsoft Live Mesh [15] aims to provide a Grid. Finally, he can execute and monitor the
centralized location for a user to store applications and application using the Sun Grid Web portal or API.
After the completion of the application, the user will provider and sub-leasing these to the consumers. A
need to download the execution results to his local broker can accept requests from many users who have
development environment for viewing. a choice of submitting their requirements to different
GRIDS Lab Aneka [17], which is being brokers. Consumers, brokers and providers are bound
commercialized through Manjrasoft, is a .NET-based to their requirements and related compensations
service-oriented platform for constructing enterprise through SLAs. An SLA specifies the details of the
Grids. It is designed to support multiple application service to be provided in terms of metrics agreed upon
models, persistence and security solutions, and by all parties, and penalties for meeting and violating
communication protocols such that the preferred the expectations, respectively.
selection can be changed at anytime without affecting Such markets can bridge disparate Clouds allowing
an existing Aneka ecosystem. To create an enterprise consumers to choose a provider that suits their
Grid, the service provider only needs to start an requirements by either executing SLAs in advance or
instance of the configurable Aneka container hosting by buying capacity on the spot. Providers can use the
required services on each selected desktop computer. markets in order to perform effective capacity
The purpose of the Aneka container is to initialize planning. A provider is equipped with a price-setting
services and acts as a single point for interaction with mechanism which sets the current price for the
the rest of the enterprise Grid. Aneka provides SLA resource based on market conditions, user demand, and
support such that the user can specify QoS current level of utilization of the resource. Pricing can
requirements such as deadline (maximum time period be either fixed or variable depending on the market
which the application needs to be completed in) and conditions. An admission-control mechanism at a
budget (maximum cost that the user is willing to pay provider’s end selects the auctions to participate in or
for meeting the deadline). The user can access the the brokers to negotiate with, based on an initial
Aneka Enterprise Grid remotely through the Gridbus estimate of the utility. The negotiation process
broker. The Gridbus broker also enables the user to proceeds until an SLA is formed or the participants
negotiate and agree upon the QoS requirements to be decide to break off. These mechanisms interface with
provided by the service provider. the resource management systems of the provider in
order to guarantee the allocation being offered or
4. Global Cloud Exchange and Markets negotiated can be reclaimed, so that SLA violations do
not occur. The resource management system also
Enterprises currently employ Cloud services in order to provides functionalities such as advance reservations
improve the scalability of their services and to deal that enable guaranteed provisioning of resource
with bursts in resource demands. However, at present, capacity.
service providers have inflexible pricing, generally
limited to flat rates or tariffs based on usage thresholds, Brokers gain their utility through the difference
and consumers are restricted to offerings from a single between the price paid by the consumers for gaining
provider at a time. Also, many providers have resource shares and that paid to the providers for
leasing their resources. Therefore, a broker has to
proprietary interfaces to their services thus restricting
choose those users whose applications can provide it
the ability of consumers to swap one provider for
maximum utility. A broker interacts with resource
another.
providers and other brokers to gain or to trade resource
For Cloud computing to mature, it is required that shares. A broker is equipped with a negotiation module
the services follow standard interfaces. This would that is informed by the current conditions of the
enable services to be commoditised and thus, would resources and the current demand to make its
pave the way for the creation of a market infrastructure decisions.
for trading in services. An example of such a market
Consumers have their own utility functions that
system, modeled on real-world exchanges, is shown in
cover factors such as deadlines, fidelity of results, and
Figure 4. The market directory allows participants to
turnaround time of applications. They are also
locate providers or consumers with the right offers.
constrained by the amount of resources that they can
Auctioneers periodically clear bids and asks received
from market participants. The banking system ensures request at any time, usually by a limited budget.
Consumers also have their own limited IT
that financial transactions pertaining to agreements
infrastructure that is generally not completely exposed
between participants are carried out.
to the Internet. Therefore, a consumer participates in
Brokers perform the same function in such a market the utility market through a resource management
as they do in real-world markets: they mediate between proxy that selects a set of brokers based on their
consumers and providers by buying capacity from the
 Compute Cloud

Storage Cloud
Broker 1

Request Negotiate/Bid Publish Offers

Capacity

Directory
.
. Bank
.
Auctioneer
Enterprise .
Resource
Manager
(Proxy) Broker N
Global Cloud Compute
Cloud
Market
Enterprise IT Consumer

Storage Cloud

Figure 4: Global Cloud exchange and market infrastructure for trading services.
offerings. He then forms SLAs with the brokers that obtain restitution in case an SLA is violated. This
bind the latter to provide the guaranteed resources. The motivates the need for a legal framework for
enterprise consumer then deploys his own environment agreements in such markets, a research issue that is out
on the leased resources or uses the provider’s of scope of themes pursued in this paper.
interfaces in order to scale his applications.
The idea of utility markets for computing resources 5. Summary and Conclusion
has been around for a long time. Recently, many
research projects such as SHARP [18], Tycoon [19], Cloud computing is a new and promising paradigm
Bellagio [20], and Shirako [21] have come up with delivering IT services as computing utilities. As
market structures for trading in resource allocations. Clouds are designed to provide services to external
These have particularly focused on trading in VM- users, providers need to be compensated for sharing
based resource slices on networked infrastructures such their resources and capabilities. In this paper, we have
as PlanetLab. As mentioned before, the Gridbus project proposed architecture for market-oriented allocation of
has created a resource broker that is able to negotiate resources within Clouds. We have discussed some
with resource providers. Thus, the technology for representative platforms for Cloud computing covering
enabling utility markets is already present and ready to the state-of-the-art. We have also presented a vision for
be deployed. the creation of global Cloud exchange for trading
services.
However, significant challenges persist in the
universal application of such markets. Enterprises The state-of-the-art Cloud technologies have limited
currently employ conservative IT strategies and are support for market-oriented resource management and
unwilling to shift from the traditional controlled they need to be extended to support: negotiation of
environments. Cloud computing uptake has only QoS between users and providers to establish SLAs;
recently begun and many systems are in the proof-of- mechanisms and algorithms for allocation of VM
concept stage. Regulatory pressures also mean that resources to meet SLAs; and manage risks associated
enterprises have to be careful about where their data with the violation of SLAs. Furthermore, interaction
gets processed, and therefore, are not able to employ protocols needs to be extended to support
Cloud services from an open market. This could be interoperability between different Cloud service
mitigated through SLAs that specify strict constraints providers.
on the location of the resources. However, another As Cloud platforms become ubiquitous, we expect
open issue is how the participants in such a market can
the need for internetworking them to create a market- [10] Morgan Stanley. Technology Trends. 12 June
oriented global Cloud exchange for trading services. 2008.
Several challenges need to be addressed to realize this http://www.morganstanley.com/institutional/tec
vision. They include: market-maker for bringing hresearch/pdfs/TechTrends062008.pdf [18 July
service providers and consumers; market registry for 2008]
publishing and discovering Cloud service providers [11] K. Keahey, I. Foster, T. Freeman, and X.
and their services; clearing house and brokers for Zhang. Virtual workspaces: Achieving quality
mapping service requests to providers who can meet of service and quality of life in the Grid.
QoS expectations; and payment management and Scientific Programming, 13(4):265-275,
accounting infrastructure for trading services. Finally, October 2005.
we need to address regulatory and legal issues, which
[12] I. Llorente, OpenNebula Project.
go beyond technical issues.
http://www.opennebula.org/ [23 July 2008]
[13] Amazon Elastic Compute Cloud (EC2),
http://www.amazon.com/ec2/ [18 July 2008]
[14] Google App Engine,
http://appengine.google.com [18 July 2008]
[15] Microsoft Live Mesh, http://www.mesh.com
[18 July 2008]
References [16] Sun network.com (Sun Grid),
[1] L. Kleinrock. A vision for the Internet. ST Journal http://www.network.com [18 July 2008]
of Research, 2(1):4-5, Nov. 2005. [17] X. Chu, K. Nadiminti, C. Jin, S. Venugopal,
[2] S. London. INSIDE TRACK: The high-tech and R. Buyya. Aneka: Next-Generation
rebels. Financial Times, 06 Sept. 2002. Enterprise Grid Platform for e-Science and e-
http://search.ft.com/nonFtArticle?id=02090600 Business Applications. In Proceedings of the
0371 [18 July 2008] 3th IEEE International Conference on e-Science
and Grid Computing (e-Science 2007),
[3] I. Foster and C. Kesselman (eds). The Grid: Bangalore, India, Dec. 2007.
Blueprint for a Future Computing
Infrastructure. Morgan Kaufmann, San [18] Y. Fu, J. Chase, B. Chun, S. Schwab, and A.
Francisco, USA, 1999. Vahdat. SHARP: an architecture for secure
resource peering. ACM SIGOPS Operating
[4] M. Chetty and R. Buyya. Weaving Computational
Systems Review, 37(5):133–148, Dec. 2003.
Grids: How Analogous Are They with Electrical
Grids? Computing in Science and Engineering, [19] K. Lai, L. Rasmusson, E. Adar, L. Zhang, and
4(4):61–71, July–Aug. 2002. B. A. Huberman. Tycoon: An implementation
[5] A. Weiss. Computing in the Clouds. netWorker, of a distributed, market-based resource
11(4):16-25, Dec. 2007. allocation system. Multiagent and Grid Systems,
1(3):169–182, 2005.
[6] Twenty Experts Define Cloud Computing,
http://cloudcomputing.sys- [20] A. AuYoung, B. Chun, A. Snoeren, and A.
con.com/read/612375_p.htm [18 July 2008]. Vahdat. Resource allocation in federated
[7] R. Buyya, D. Abramson, and S. Venugopal. The
distributed computing infrastructures. In
Grid Economy. Proceedings of the IEEE, 93(3): Proceedings of the 1st Workshop on Operating
698-714, IEEE Press, USA, March 2005. System and Architectural Support for the
Ondemand IT Infrastructure (OASIS 2004),
[8] S. Venugopal, X. Chu, and R. Buyya. A
Boston, USA, Oct. 2004.
Negotiation Mechanism for Advance Resource
Reservation using the Alternate Offers Protocol. In [21] D. E. Irwin, J. S. Chase, L. E. Grit, A. R.
Proceedings of the 16th International Workshop Yumerefendi, D. Becker, and K. Yocum.
on Quality of Service (IWQoS 2008), Twente, The Sharing networked resources with brokered
Netherlands, June 2008. leases. In Proceedings of the 2006 USENIX
[9] D. Hamilton. 'Cloud computing' seen as next Annual Technical Conference (USENIX 2006),
wave for technology investors. Financial Post, Boston, USA, June 2006.
04 June 2008.
http://www.financialpost.com/money/story.html
?id=562877 [18 July 2008]