Cloud Provider Evaluation
Cloud Computing Adoption for SMEs: Challenges, Barriers and Outcomes by Mojtaba Akbari from ARROW@DIT is available under a Creative Commons Attribution-NonCommercial-
ShareAlike 3.0 Unported license. UMGC has modified this work and it is available under the original license.
19
3 MIGRATION OF SMES TO CLOUD
3.1 Introduction
Information Technology has become a crucially important part of the today’s business
world impacting almost all industries and sectors. Small and Medium Enterprises
(SMEs) on the other hand are considered as a vital element of the global economy.
Due to the specific requirements and characteristics of smaller organizations (such as
number of employees and budget) compared to larger corporates, the attitude of the
SMEs towards new IT technologies are normally more conservative than that of their
bigger counterparts. This chapter provides a definition of an SME, and describes the
characteristics that distinguish them from large enterprises. Then existing research
studies addressing the factors that influence adoption of new IT technologies by SMEs
are reviewed. Subsequently, the chapter focuses on cloud technology, and reviews
existing research studies that have been specifically discussing adaptation of cloud
computing by SMEs.
3.2 Definition of Sm all and Medium E nterprises
Small and Medium Enterprise (SME), also known as Small and Medium Business
(SMB), is a term used to describe companies with the number of employees below a
certain threshold. While the term ‘SME’ is more common within the European Union
and is used by major international organizations, such as, the United Nations (UN) and
the World Bank, SMB is more frequently used in United States (Georgios et al., 2001).
Until very recently, the definition of SME was different amongst various European
countries. For example, in Germany any company with a total headcount of less than
255 people was considered as an SME while this limit was 100 in Belgium or Greece
for example (Georgios et al., 2001). In 2011, the European Commission decided to
adopt a unified definition for SMEs. Within this perspective, SMEs are classified to
three categories of 1) micro-entities that are companies with less than 10 employees,
2) small companies with up to 50 headcounts and finally 3) medium-sized companies
are considered those which less than 250 permanent employees. From the financial
perspective, an SME can be defined as an enterprise with the revenue of €10-50
million. The definition is not very straightforward in US as it depends on different
criteria, such as, the industry, ownership structure and revenue.
20
3.3 Characteristics of SMEs
The traditional view of SMEs was that they are in fact the same as the bigger
companies with only difference in size. However, it is currently a well-known fact that
the very size constraints of the SMEs make them distinct from their larger
counterparts. Understanding of such characteristics and distinctions are especially
important when considering IT adoption strategies.
As a result of their small size, the major differentiator between an SME and a large
enterprise is a level of available resources to the SME, also known as “the SME’s
resources poverty” (Welsh and White, 1981). Many different factors contribute to the
resource poverty of SMEs. For example, SMEs are normally clustered in highly
fragmented industries, such as, retailing and services where there exist many
competitors. Many of these competitors deploy aggressive price-cutting strategies in
order to build revenue and to survive.
In addition, most SMEs have an owner-manager model in that a large fraction of the
overall profit is considered as the owner’s salary leaving little for future investments.
Furthermore, external environment changes tend to have a more significant impact on
the SMEs compared to their larger counterparts. For example, changes in the tax laws
or interest rates are likely to have a more significant impact on the overall expenses of
an SME compared to a bigger company.
All these factors result in a short-range management perspective where liquidity is a
main objective. As a result, SMEs may be very conservative in adopting new IT
solutions especially during its early days.
3.4 IT adoption strategies by SMEs
This section provides an overview of the earlier research studies that focus on
strategies of SMEs in adopting new IT solutions. Various literature provide multiple
definitions for IT. For example, according to Hollander et al., IT refers to the
technology side of Information Systems (Hollander et al., 2000). Similarly, Laudon
and Laudon defined the IT as any technology that enable computer based information
systems (Laudon and Laudon, 2000). However, following the work of Sarosa and
Zowghi, this study adopts a more generic definition of IT that includes any technology
that is deployed by an organization to collect, process and disseminate information in
21
all its form (Sarosa and Zowghi, 2003). Hence, this definition covers various elements
of IT including hardware, software and communication infrastructures. Furthermore,
the term “IT adoption” considered in this study refers to two concepts based on two
different but related definitions. According to Thong and Yap, IT adoption is defined
as deploying IT to support operations, management, and decision making within the
organization (Thong and Yap, 1996). Another definition is given by Jaakkola, in that
“IT adoption” is considered as replacing a traditional and old IT solution (or a non-IT
based solution) with a more advanced one, to achieve the same business goals within
the organization (Jaakkola, 1996). This study covers both definitions.
Fink, D. has conducted a research study to identify factors that impact the effective
adoption of new IT trends by SMEs. As a result of that Fink has identified three major
IT acquisition phases and their significance. Accordingly, in the first phase, SMEs
need to be able to assess a given IT solution and its benefits for their organization
considering the organization culture and employees. Secondly, they must decide
whether sufficient internal resources are available for implementation of that particular
IT solution. Finally, SMEs need to be able to evaluate external environment support
and resources, especially if internal in-house resources are lacking. At the end of each
phase, the SME should make decision whether or not it should continue considering
that particular IT solution (Fink, 1998).
Similarly, Thong has examined the relationship between a number of contextual
factors, including the CEO innovativeness, IT knowledge and his/her attitude towards
new IT technologies, and IT adoption in SME organizations (Thong and Yap, 1996).
As a result, the study conclude that while the adoption of new IT solution is expectedly
dependent on the innovativeness and attitude of the CEO towards new technologies,
the extent of IT adoption is merely determined by CEO’s level of IT knowledge.
Syed Alam, has examined the impact of five major factors in adopting new IT
solutions by SMEs in two states of Malaysia (Alam, S., 2009). The five investigated
factors are 1) perceived benefits 2) perceived cost 3) ICT knowledge and skill 4)
external pressures and 5) government supports. The researcher concluded that the
perceived benefits of an ICT solution have greatest impact of the adoption decision.
22
In a similar study, Seyal et al. have conducted a survey of fifty four SMEs to
investigate the major factors impacting the use of IT within those organizations (Seyal
et al., 2000). As a result of the study, the authors have identified three major
parameters, namely: the sales figures of the organization, the characteristics of the
business and the CEO’s computer knowledge and skill level. As a result, the study
recommends that CEOs keep themselves updated with the IT technology and to
additionally encourage the use of IT wherever applicable within the organization,
including not only the management functionalities but additionally the operational
functions too.
While there exists many other similar, and to some extent overlapping studies, one can
broadly classify all the impacting factors to two categories namely internal and
external drivers and barriers for adopting IT solutions. As mentioned by several
previous studies, major internal factors are 1) organizations’ CEO 2) employees and 3)
available resources. On the other hand, most influential external factors are 1)
competitors 2) government 3) suppliers and customers and finally 5) vendors and
consultants (Seyal et al., 2000).
3 . 4 . 1 I n t e r n a l f a c t o r s i m p a c t i n g a n o r g a n i z a t i o n ’ s I C T
The role of CEO
In most SME environments, the owner is also the manager of the company.
Therefore, such a person needs to perform all managerial functions including
planning for
business road map, organising resources including staff, evaluation and executing
decisions, as well as, monitoring the functionalities of all parts of the organization.
23
Considering this intensive workload, many studies argue that the manager can benefit
from the support of IT technologies (Cragg and King 1993; Fink 2002; Lanz 2002;
Mehrtens et al. 2001; Mirchandani and Motwani 2001).
In fact, to many researchers, the perceived benefit of IT by the owner can be
considered as a deciding factor for whether or not to deploy a given IT solution.
Accordingly, the limited knowledge of the owner can be considered as a major barrier.
In addition to that, considering the fast evolution of the IT industry, the owners of
SMEs may be confused with regards to the choices of available IT solutions which
may have negative impact towards IT adoption in general (Venkatesh and Brown
2001). The other extreme case in this spectrum is when the owner has an adequate
understanding of a given IT solution which may discourage him from deploying that
solution due to the concerns regarding the potential shortcomings of that given
technology/strategy (Agarwal and Prasad , 2000; Love et al., 2001).
Furthermore, Vincent, have identified the openness of an SME’s owner as another
decisive factor (Vincent and Guynes, 1997). According to Rogers, openness can be
defined as a degree to which the owner is connected to people outside the organization
(Rogers, 1983). According to Saleh and Wang, such openness can results in increased
awareness of the owner in respect to the technology which ultimately can have an
encouraging impact on his/her decisions regarding IT deployment strategies (Saleh and
Wang, 1993). Finally, Thong and Yap, have argued that the subjective norm, defined
as the perceived social pressure to engage or not to engage in behaviour, is additionally
important when it comes to IT adoption (Thong and Yap, 1995). To summarise, the
key factors that have significant impact on adaptation of IT deployments in regards to
the companies’ manager/owners are identified as 1) Perceived benefits of the given IT
solution 2) Computer literacy of the CEO 3) Assertiveness and openness of the CEO 4)
Perceived control of the CEO and finally 5) Subjective norms.
Resources
According to Chau, Utomo and Dodgson, availability of resources, and especially
financial resources, is extremely important when considering success of a given IT
solution/product adaptation in an SME organization (Chau, 1995; Utomo and
Dodgson, 2001). This is mainly due to the limitation of SME’s financial resources
which makes them very vulnerable to the short term fluctuations of every aspect of the
24
business (Welsh and White, 1981). Hence, liquidity is the foremost concern of all
SMEs and this conservativeness manifests itself in the effort of SMEs to prioritise their
IT needs and subsequently their deployment strategy. According to Sarosa and Zowghi
and Schollhammer, there exist numerous examples of inappropriate IT deployment
decisions that have led to the bankruptcy of whole organizations (Sarosa and Zowghi,
2003; Schollhammer and Kuriloff, 1979). Consequently, Agarwal, Prasad and Love et
al., have argued that despite of all potential perceived benefits, some CEOs may be
discouraged from deploying costly IT solutions because of their high risk and financial
impact on the organization (Agarwal and Prasad, 2000; Love et al., 2001).
As a result of the limited financial resources available to most SMEs, it may be
additionally more difficult to obtain necessary external expertise or additional training
from vendors or IT providers. Attewell, has identified this as a primary factor that
leads SMEs to delay the in-house adoption of complex new technologies in IT
domains until the technology becomes more mature and SMEs obtain sufficient
technical knowledge deployment and maintenance of that technology (Attewell, 1992).
To lower the knowledge barrier, many organizations have been formed with the
objective of simplifying technology adoption for small firms. As an example of such
entities, mention can be made to Service Bureaus Consultancy firms. In line with this
vision, Utomo and Dodgson, have shown that the SMEs are more eager to adopt new
IT solutions when government assistance or voluntary consultancy from higher
education institutions, is available (Utomo and Dodgson, 2001).
The associated high risk of IT investment could discourage some owners/managers to
adopt IT for their company (Agarwal and Prasad, 2000; Love et al., 2001). Limited
financial resources also mean that SMEs might not be able to obtain necessary external
expertise, such as, consultant or additional training from vendors (Attewell, 1992).
Employees
In scenarios where employees are the users of a given IT solution, their acceptance and
openness towards that solution/product can influence the decision of the CEO.
According to Davis, users are more likely to support adoption of a new IT solution if
they can understand its positive impact on their own aspect of the business and feel
comfortable using that technology (Davis, 1989). Additionally, according to Attewell
25
and Love et al., employees will be considerably more supportive of a new IT
technology, if relevant trainings are made available (Attewell, 1992; Love et al., 2001).
However, as discussed earlier, this may not be possible in many scenarios, due to the
limited financial resources available for most SMEs.
Furthermore, employees may be reluctant to adopt a new IT technology due to various
reasons, such as, the effect on their routine work practice and their lack of self-steam in
terms of capabilities to learn and adopt that new system (Drew, 2003; Love et al.,
2001). In this regard, the CEO has the power to influence, motivate, or even force the
employee to adopt the new technology.
3 . 4 . 2 E x t er n a l f a ct o r s i m p a ct i n g a n o r g a n i z a t i o n ’ s I C T
Competitors
Considering the high level of competition that exists in most SMEs’ area of business,
deployment of an appropriate IT technology can be considered as a differentiator (Earl,
1989; Galliers and Sutherland 1999; Turban et al. 2002). According to Galliers and
Sutherland a new IT technology can affect the competitive forces that shape an
industry by (Galliers and Sutherland, 1999):
i. Building barriers against new entrants
ii. Changing the basis of competition
iii. Changing the balance of power in supplier relationships
iv. Tying in customers
v. Switching costs, and
vi. Creating new products and services
Suppliers and customers
Efficient IT solutions can provide significant gains for enterprises in regards to their
connection to both suppliers and customers. According to Ronchi, the consequence of
the introduction of new IT technologies is rather complex and, although they do not
change dramatically traditional business concepts, their influence on supply chain
management and companies’ relationships is significant (Ronchi, 2003). For example,
introduction of the Internet has changed the nature of traditional relationships, and has
led to many new “avenues of revenue” for many enterprises. According to Hollander et
al., SMEs can befit from IT adoption by (Hollander et al., 2000):
26
Improving the efficiency of their transactions processing
Reducing the overhead and costs of transactions
Improving the response time to the suppliers and consumers
Improving the marketing strategies to costumers
However, interestingly, the research by Sarosa and Zowghi suggests that the choice of
IT solution being deployed by an SME is not influenced by compatibility of that IT
solution to the customers’ system (Sarosa and Zowghi, 2003). In other words, SMEs
are less likely to adopt a particular IT solution merely to be able to integrate more
easily with their customers. Of course, this observation will additionally depend on the
nature of the business and the environment as well.
Government
As discussed earlier, due to the limitation of financial and human resources, SMEs
may need help from external sources, such as, the government, in order to deploy an IT
solution. This is especially the case when the core business of the SME is not IT
related (Sarosa and Zowghi, 2003). However, in many developed countries, there is a
gap between the level of support provided by the government and the actual need of
SMEs when evaluating or implementing an IT solution (Dutta and Evrard, 1999).
Furthermore, some governments prefer to deploy a welfare model where they hand out
the assistance package directly to SMEs (Dutta and Evrard 1999), and others choose
to formulate policy and provide access to enabler infrastructure to allow SMEs
development (Walczuch et al., 2000).
3.5 Cloud com puting for SMEs
As discussed earlier, limitations in the availability of financial and human resources
may delay SMEs in adopting new technologies, specifically, during its early stages.
This is especially true considering less complex IT requirements of SMEs compared to
their larger counter parts. On the other hand, the studies of Bose and Vijeikis reveal
that SMEs are normally interested in outsourcing their IT requirements in order to be
able to focus more on their own core business space (Bose and Sugumaran, 2006;
Vijeikis and Makstutis, 2009). This could make cloud computing an attractive option
for SMEs. The research studies of Donelaicio and Taylor summarise the advantages
and disadvantages of cloud solutions for SMEs (Donelaicio, 2012; Taylor et al., 2010).
27
Accordingly, the advantages can be listed as 1) Cost reduction 2) Respond to business
requirements 3) Easier administration and maintenance and 4) Global access.
Cost reduction
Depending on the type of cloud services being deployed, significant cost reduction can
be achieved in hardware and software provisioning. In addition, the overall
management and maintenance costs may also be reduced. In other words, while cloud
computing is normally envisioned as a way of converting capital expenditure
(CAPEX) to operational expenditure (OPEX), a successful migration strategy may also
reduce the OPEX (Armbrust et al., 2010). Financial gains can be main drivers for
migration of an organization to cloud solutions which are even more critical when
SMEs are considered. This is confirmed by the IDC survey 2009 which shows the
prime reason behind the adoption of cloud computing is cost-saving, according to 50%
of the respondents (IDC, 2009). In another survey study by Cisco, 512 SMEs who
have considered migrating to cloud or who are already using cloud services are
surveyed. Interestingly, on average, they rated their level of satisfaction as 7.8 (out of
10), with less than 10 per cent of respondents rating their satisfaction as 5 or less
(Taylor et al., 2010). The biggest sources of value from their current cloud services
come from maximizing investment, followed by the fact that they could focus on their
business rather than on technology. As an example of a case study evaluating the
financial gains of migrating to cloud, mention can be made to the work by Khajeh-
Hosseini et al. which shows a 37% cost saving result over 5 years when moving from
an in-house data centre to a cloud infrastructure (Khajeh-Hosseini et al., 2010).
According to Ambrust et al., the decision of migrating to cloud should be made if the
average cost of the system after migration is substantially lower than using in-house
solutions and that migration process does not imply a considerable burden on the
overall return of investment (Ambrust et al., 2009). The latter case is of special
importance for SMEs.
Respond to business requirements
Migrating to the cloud system may significantly simplify the software upgrades, as
well as, adding a new software application. Normally, the cloud providers upgrade
their software applications more frequently than small enterprises can afford to.
Additionally, after migration to cloud, it will be significantly easier and faster to
expand the business and resources.
28
Easier administration and maintenance
Cloud solutions offer significantly easier system administration and maintenance.
Users may only need web browsers or other simple client applications to have access
to a diverse range of applications. Additionally, the software consistency between the
users is guaranteed and the administrator can be assured that all users use the same
version of applications. Furthermore, the data backup can be realised in a near real-
time fashion which lowers the risk data loss.
Global access
Using a cloud solution, employees, partners and clients can easily access to the
information and applications and update necessary information regardless of their
physical location. This flexibility is especially important for high mobility employees.
On the other hand, major risks of migration to a cloud-based solution can be listed as
1) Security and privacy concerns 2) Connectivity and performance and 3) Control loss
and dependency.
Security and privacy concerns
Security and privacy concerns are considered as one of the major barriers in adopting
cloud computing. For example, Cisco survey findings reveal that respondents consider
security as the most important factor when choosing a cloud provider (Taylor et al.,
2010). More specifically speaking, the survey defines the level of importance of
various factors that is quantity, varying from 1 (least important) to 10 (most
important). Based on surveying 512 correspondents, security was identified as the
most important factor, with the average level of importance equalling 7.8. Similarly,
based on the report results, nearly 40% of respondents believe that interest in buying
cloud services will be significantly increased if security and privacy can be assured.
The study of information security of cloud computing for enterprises (Tsai and Lin,
2011) provides a comprehensive overview of security related issues and potential
solutions for cloud environments.
Connectivity and performance
Since all cloud services are based on the network infrastructure, failure of the network
system may lead to unavailability of all dependent services and applications. In
addition, despite significant growth of communication networks and successful
29
penetration of high speed broadband solutions in recent years, the end-to-end delay and
network congestion may still be of concern for many SMEs (Tsai and Lin, 2011).
Control loss and dependency
SMEs may be afraid of losing control over the management of their IT infrastructure
and become dependent to cloud providers. Concerns will further be amplified if they
need to commit to long term contracts or feel that migration to another provider may
not be a straightforward process (Tsai and Lin, 2011).
3.6 Conclusion
After identifying characterises of SMEs in general, this chapter has reviewed the
research literature addressing the important factors for SME organizations in the
adoption of new IT technologies in general. As a result, a number of key factors have
been identified as commonly mentioned in different research studies. Such factors
have been classified to two categories, namely internal and external factors. The role
of the CEO, available resources and the support of employees can be mentioned as the
most important internal factors. In addition, there also exist a number of external
factors such as the role of consumers and the government. The chapter then focused on
cloud computing in particular and reviewed research studies that have addressed
various aspects of cloud computing technology in relation to SMEs.
Through deployment of the real cloud-based pilot system, this study aim to collect
feedback from SMEs’ employees who will be using the pilot system during the
evaluation period. Additionally, two survey studies will be conducted to complement
this experimental work. The next chapter explains the steps involved in the design and
implementation of the pilot system, as well as, the rationale behind the survey
questions.
30
4 DESIGN AND DEVELOPME NT
4.1 Introduction
This chapter summarises the main aspects involved in the design and development of
research methodologies used in the study. More specifically speaking, the chapter is
divided to three major sections as follows: Section 4.2, gives overview of the
objectives of the experimental set up used in this thesis. It explains the cloud
architecture and the related performance metrics. Subsequently, Section 4.3 describes
the steps involved in implementation of the design and configuration discussed earlier
in Section 4.2. Finally, Section 4.4 focuses on the main objectives of conducting the
survey studies. It explains the two categories of surveys used through this study and
reasons behind such classifications.
4.2 Design overview
This section summarises the design aspect of the experimental set up used through this
study. More specifically, the participating SME organizations are introduced and the
reason behind choosing those enterprises for this study are explained. Subsequently,
the choice of the cloud provider, as well as, the services used for evaluation in this
study are explained. The chapter further explains the cloud architecture and the related
performance metrics.
4 . 2 . 1 E x p e r i m en t a l p i l o t s ys t em
To evaluate advantages and disadvantages of potential migration to cloud by SMEs, a
real implementation has been desired. It is critical to observe the performance of a real
cloud solution for a number of services including cloud-based storage.
The experiment design aims to provide a vision and demonstration of a real cloud
migration from an SME perspective. Three SMEs from completely different business
domains were selected and different type of cloud implementation has been examined
for each of them. The evaluation of the pilot system includes qualitative feedback from
those users who have been using the services, as well as, analytical measurements and
demonstrations. The qualitative feedback focuses mainly on users’ quality of
experience in terms of latency and reliability of the system (availability) compared to
the existing IT platform. Subsequently the quantitative evaluation consists of two
31
demonstrations, one showing the impact of access latency, and the other one address
the issue of computation latency. Details of the evaluation methodologies are further
explained in the next chapter. The following provides an overview of the enterprises
considered in this study.
1. O’Brien PR
O’Brien PR is a service agency with a main focus of providing its clients services
through offering an individual, personalised and professional service towards their
public relations needs. A wide range of communication solutions are being offered by
this very small-sized enterprise. The organization is also member of the Public
Relations Institute of Ireland (PRII).
O’Brien PR’s focus is on designing communications and public relations strategies to
ensure the client’s message reaches the right audience at the right time. This key
contribution requires a medium level of IT provisioning. The enterprise is very
dependent on its electronic and computerized communication and operations. And
while the business is a very small organization, the aim of choosing this company as
one of the pilot participant businesses is to explore the tendency of a small owner-
manager organization adopting cloud computing.
Since the aim is to investigate real owner-manager behaviour, due to the current
server-client setup at the organization, and also the simplicity of the type of services
being used with no specialized application requirements, the migration is expected to
be straight forward.
Most organizations who have recently implemented an upgrade on their traditional
infrastructure are no longer keen to upgrade again as they would see this as huge waste
of money and resources. The final reason O’Brien PR was selected as part of this
research was the out-dated status and poor perfuming quality of their traditional
infrastructure. Therefore, O’Brien PR were keen to learn about new cost effective
solutions based on cloud computing.
32
2. Translation.ie
As the name implies, translation.ie is a translation and interpreting organization that
was established in 2000. The translation of various documents, from almost every
language to English and from English to many world languages, is one of the primary
services provided by the organization. The service covers a wide range of translation
areas, such as, technical, legal, medical, financial and engineering. Additionally, the
site offers website and software localisation which removes cultural barriers among
international trading (clients/organizations).
Since the company deals with data from extremely sensitive clients, such as, Court-
Service Ireland, An Garda Siochana, the Health Service Executive, the Department of
Foreign Affairs and even the office of the Revenue Commissioners, privacy and
security is a vital matter when it comes to migrating to cloud. Adoption of cloud
computing for translation.ie has always been bringing the concerns around security and
privacy.
Based on the interactions with the company, it is understood that while the company is
concerned about the security and privacy aspect, it is simultaneously looking for
potential solutions that allow flexible access and the sharing of work files among their
staff. Subsequently, the organization was approached to examine how well the cloud-
based pilot system can satisfy their needs and requirements.
3. The K Club
The K Club is a country club, located in the heart of the idyllic Irish countryside in the
pretty village of Straffan, Co. Kildare. The resort comprises accommodation,
restaurants, retail outlets, a spa and golf facilities, each of which has its own
customized software. The organization has around 40 applications, some as simple as
Exchange, Active Directory, payroll and GPS tracking system or as complex as
Property Management System (PMS) which is the organization’s main business
application. There are already some cloud-based applications in use, such as, SynXis
(the Central Reservation System) integrated with the in-house PMS, providing a
holistic view of their revenue.
33
With 120 users across five fibre-linked buildings and with difficulties and expenses
involved in managing ICT for an organization whose core business is not related to IT,
The K Club, as a medium sized organization, presents an ideal candidate for the
purpose of this pilot study. Furthermore, the hotel industry is all about flexibility
therefore, building a flexible IT should bring a lot of comfort to the business. It would
be interesting to measure/demonstrate the flexibility of the proposed cloud-based pilot
solution.
Another reason for The K Club’s suitability, was due to firstly it being a good example
of a medium sized enterprise in the hospitality sector (as discussed above), and
secondly, because of the personal affiliation of the author to the business (as head of
the IT department). Due to this access, it was quite possible to evaluate different
configurations while making sure that service interruption was avoided/minimized.
Moreover, the location of The K Club in the Irish countryside gives the opportunity to
study the challenges that an organization located in a sub urban/rural area may face in
terms of broadband connectivity; an essential part of most cloud-based solution.
Interestingly, there have been a number of recent studies addressing the adoption of
ICT within the hospitality and tourism sector in Ireland (Duffy, 2010).
4 . 2 . 2 C h o i c e o f v e n d o r
SME’s current technology services providers are the best choice to get deals on cloud
services, however, some current suppliers for the SMEs named in 4.2.1, were
approached, but no cloud offering was available for the purpose of this study.
Examples of such service providers are Eircom, UPC and Vodafone.
For the purpose of the pilot implementation, a number of key cloud players in Ireland
have been studied and contacted to discuss possible implementation opportunities as
part of the research study.
34
The main criteria for choosing a vendor for research purposes was based
on availability, estimated cost and type of services which would be offered. To
choose the most suitable one, a limited number of cloud providers were
shortlisted and contacted to understand about their available services and SLAs.
An important decision regarding the cloud migration is choice of cloud provider. This
is especially true considering the vast range of cloud services providers and offers
available to organizations. As Figure 4.1, illustrates, there are a lot of local and
international cloud providers in Ireland for various aspects of cloud
computing services. Some key providers for virtualization can be named as
VMware, Redhat, Citrix, Oracle and Microsoft but when it comes to Infrastructure
as a Service, other international providers, such as, HP, EMC, Dell, Rackspace, IBM
and CA play a major role. From the SaaS perspective also, the availability of cloud
services depends on the business sector and hence cloud providers are more
specialized. For example the SaaS providers focusing on financial service
applications are different than those who provide SaaS for logistics or retail
applications.
35
IBM has been very active in many aspects of cloud computing, in Ireland, both in
research and development and also in strategic hosting. The company established its
first cloud computing centre in Ireland in 2008.
With Ireland hosting 9 of the top technology companies, who offer cloud services, the
country has a good supply of commercial operators offering co-location and managed
services. These include BT, Cable and Wireless, Verizon, HP, Eircom, Servecentric,
Data Electronics, Interxion, Digiweb, CIX (Cork Internet Exchange) and some other
international companies, such as, Microsoft, IBM, HP, Google, Dell and EMC.
IBM was considered as a cloud provider in this study as firstly, it is very well
established in Ireland and secondly it provides a wide range of cloud services publicly
that have enabled the pilot infrastructure study to be established on their public cloud
system. Other vendors, such as, VMware, EMC and Sungard were also approached
however the SLAs and cost factors have seen to be more beneficial with IBM.
4 . 2 . 3 C l o u d o f f er i n g s er v i c es
There were two cloud offerings considered for the purpose of this study’s pilot system.
The pilot is focused on choosing services to demonstrate a live example of SaaS and
IaaS. As will be further discussed, platform offering as a service (PaaS) is not
considered in this study.
Both services will be hosted on IBM’s smart cloud services which (public cloud IaaS),
and provide a provisioned and scaled environment including storage, network, and
middleware and computing. IBM’s data centres which have required services for the
purpose of the pilot system, are available globally, however, the datacentres in
Ehningen, Germany and Frankfurt, Germany have been considered.
IBM provides other cloud services, such as, platform services and private cloud
offerings. The pilot system however, will not use such offerings.
Object Storage offering of IBM (Storage IaaS)
Object Storage is a pure storage offering. It is suggested that Object Storage is a Web-
API based file system suitable for unstructured data and has no limit in terms of
capacity. The goal of using this service is to measure scalability and understand the
challenges of storage growth. The idea is to allow uploading, storing and organizing
36
any type of file from anywhere in the world regardless of the file size and to be able to
access and update that data globally.
This service is an ideal solution for all three participating organizations. The reason for
this is because the files that are used for the pilot study have an unstructured nature and
are associated with different applications, such as, Microsoft office spread sheets,
power-point, documents, audio, visual files, media files, email and PDF files. It would
be interesting to experiment and measure challenges of data being read and written
using the WAN infrastructure that provides connectivity to the cloud pilot
environment.
The ideal scenario for the purpose of proving storage as IaaS, is to allow participant
organizations have secure non-mediated and shared access to a centrally-managed
storage platform where space allocation, access control and security performance is
managed by the host. This is similar to in-house Storage Area Network (SAN)
architecture, which provides some data servers access to a share storage area over the
storage network. Additionally, by using an off-site storage service, an enterprise does
not need to be concerned about the disk availability, the management of the storage
system and the pay per use pricing, as this will assure the organization that resource
over-provisioning will be avoided.
The Object Storage setup will additionally allow comparison between in-house
datacentre implementation and an off-site solution. In the next chapter, a set of demos
are shown to describe the trade-offs between the two architectures.
Persistent Storage offering of IBM (IaaS/SaaS)
There are various methods of storing data objects on disks and being able to retrieve
them at a later stage, one of which is Persistent Storage. Considering a scenario of
saving a simple ‘to do’ list. The options vary from using a standard formatted text file
or CSV file or even using a rational database, such as, MYSQL, POSTGRES and DB2.
What all these methods have in common is the way data storage is separated from the
actual object programs that operate the data.
Persistent Storage can be described as an infrastructure offering that is not only
capable of managing unstructured data, but also handles structured storage, such as,
37
SQL databases, and also, offers the ability of using the storage to try Software as a
Service solution in public cloud. In the study, this later case is realized by using
remote desktop protocol. For the purpose of the pilot study, a number of virtual
machines will be deployed. Persistent Storage will be then linked to those virtual
machines. The benefit is that if the virtual systems crash or are removed, Persistent
Storage linked to that particular machine, will be still available even after that VM
removal or crash. Like any virtual scenario, Persistent Storage cannot be attached
concurrently to more than one virtual machine. There is a limitation of 2 terabyte
capacity per each block of Persistent Storage. Fortunately this is not problematic for
the purpose of this study’s experiment. While Persistent Storage supports different
operation systems, due to the requirement of SMEs considered in this study, only
Microsoft Windows Operating Systems were deployed.
Unfortunately the main drawback in terms of performance review is that the speed of
the Wide Area Network will be highly affecting the overall access time and
input/output operations of the files. This has been further investigated in the next
chapter.
4 . 2 . 4 C h o i c e o f s er v i c es t o e va l u a t e
An argument put forward by Arnold is that IaaS is the most common service and that
storage is considered to be the main service that cloud computing providers can offer
(Arnold, 2010). Additionally, many companies may benefit from using applications in
cloud (SaaS). However, when it comes to PaaS, it would be very difficult to choose a
non IT enterprise that requires PaaS and subsequently Platform as a Service has not
been considered in this study. The types of services that will be evaluated for the
participating organizations are outlined below.
O’Brien PR cloud services:
O’Brien PR was previously experiencing technical difficulties and as a result increased
service downtime. The current server infrastructure of the company is based on a 2003
windows server which is out-dated. The downtimes were mainly due to the outdated
infrastructure within the enterprise and the lack of technical expertise to resolve
technical IT issues. Consequently, the firm has been very keen on migrating to cloud,
and therefore the in-house fileserver was agreed to be replicated and migrated to cloud
for this study.
38
It is a requirement for O’Brien PR that the Wide Area network become accessible,
both with client machines and some other systems, such as, a network scanning
system, which should automatically store scanned documents online and carry out
automatic authentication without any manual intervention. Type of files for O’Brien
PR will be mainly media and financial data, such as, photos, video footage and other
frequently accessed unstructured contents. O’Brien PR would mainly use applications,
such as, Adobe Suite, Microsoft Office and Microsoft Exchange Anywhere which will
be installed on the pilot platform for two users to be used over the Wide Area Network
through remote desktop protocol (RDP).
Translation.ie Cloud Services
Due to privacy and security concerns translation.ie has shown to be more conservative
in terms of migrating their materials to cloud. However, it was agreed to move one of
their databases to the pilot system in order for two of users to share the migrated data.
The K Club Cloud Services:
Currently, The K Club has an in-house virtualised data centre consisting of a live
cluster and DR Cluster. The live cluster has two ESX machines hosting a number of
virtual instances. The live cluster is replicating with the DR cluster using VMware
services. The K Club has agreed to move a number of its storage services where the
access requirement is on a daily basis. The total number of participating users in this
study is 6 people from various departments such as sales, retail and marketing. The file
server offering will be only examined as a portion of the databases approved to be
migrated for the purpose of this study.
4.3 Im plem entation and Developm ent
This section describes the steps involved in implementing the design and configuration
discussed earlier in Section4.2. In particular, it explains how the setup is performed
along with difficulties and challenges experienced both from a technical and non-
technical perspective.
For the purpose of achieving the objectives and metrics, the evaluation process is
divided into the two categories of, quantitative and qualitative benchmarking, where
further details will be provided in Chapter 5. In summary, qualitative evaluation will
be based on face to face interviews along with some questions, to be discussed, with
39
the users in the participating organizations, in order to learn about their experience
with the deployed prototype. This is done mainly because the quantitative evaluation
cannot capture many aspects of user experience, and many enterprises have concerns
in relation to cloud deployment. The quantitative evaluation consists of two demos. In
one set of experiments, the pilot cloud-based storage is compared to the in-house data
centre, as well as, a local disk in terms of file access latency. For that purpose, a
Matlab™ based application has been developed that requires frequent I/O operations.
In the other set of experiments, the advantages of migrating into cloud, in terms of,
availability of increased computational resources are shown. For this scenario, again a
Matlab™ based application is developed in that computation is the main burden in
comparison to the communication overheads.
4 . 3 . 1 I m p l em en t a t i o n o f t h e p i l o t i n f r a s t r u ct u r e i n C l o u d
As discussed, IBM’s smart cloud enterprise is used for the purpose of the pilot system
implementation. The persistent data storage centre in Ehningen, Germany is used as
primary location for installing the virtual machines for the organizations. Other
available locations can be named as 1) Raleigh, North Carolina, United States of
America 2) Boulder, Colorado, United States of America 3) Markham, Canada 4)
Makuhari, Japan and finally 5) Singapore , Singapore .
Figure 4.2: IBM's Database locations
Consequently the object data storage centre in Frankfurt, Germany was used as
primary location for storing the organizations’ data. Other possibilities included:
Secaucus, New Jersey, United States of America
Los Angeles, California, United States of America (for storage metadata only)
40
This decision was made mainly to keep data in Europe and in order to minimize the
data access latency. The deployed platform was provisioned dynamically to deliver the
required storage capacity from the pilot system infrastructure over the Internet.
Figure 4.3: Deployed unlimited Object Storage for each participating SME
The pricing model for the offered services was based on a pay-as-you-go model that
includes storage capacity, Virtual Private Network (VPN), VMware and operating
system licences licencing. An example of a pricing model for the services offered by
IBM is shown in Figure 4.4:
Figure 4.4: Pilot system pricing for each virtual machine
As part of the linking of cloud infrastructure to organizations’ in-house infrastructure,
a number of static IP-addresses were purchased through the pay as you go model. A
short overview of installation process is outlined below:
Provisioning a static IP address
Provisioning a Persistent Storage and Object Storage
Creating and configuring virtual machine instances for each organization
41
Linking storage to the private network of the organizations
Migrating data from traditional infrastructure of organizations to the pilot
infrastructure.
Install relevant monitoring tools of performance
Provide authentication and access to each organization
One of the difficulties encountered during installation of virtual machines was the fact
that upgrading virtual machines in terms of allocated hardware resources (e.g. RAM
memory or HDD allocated to each VM) was not a straightforward task. For example if
a virtual machine is created with a particular amount of allocated memory, later
modification of the VM, by simply increasing allocated resources was not possible. In
this case, instead of upgrading the virtual machine a new instance of virtual machine
ought to be created. While this limits the flexibility of the resource allocation to the
virtual machines, it is believed that such limitation is not fundamental but rather
depends on particular configuration and implementation of vendor specific product.
Provisioning a static IP address
This was a relatively straight forward task achieved by using the console interface
provided by vendor. The reason static IP addresses were acquired was to enable remote
access from anywhere, and to additionally simplify the security policies applied to the
organizations. (Adding the IP addresses to the trusted list of firewalls within the
organizations). The list of allocated IP addresses to each organization are illustrated in
Figure 4.5: .
Figure 4.5: Pilot system virtual machines' public IP details
42
Provisioning Persistent Storage and Object Storage
Persistent Storage, as discussed, is used for both structured and unstructured data and
the configuration procedure was followed through the same interface provided by IBM
shown in Figure 4.6. The figure shows amount of storage capacity reserved for each
organization. However, as illustrated Persistent Storage is attached to the virtual
machine for one organization (translation.ie) as the other two organizations will use
Object Storage.
Figure 4.6: Pilot system Persistent Storage attachment
Moreover, the Object Storage, which is primarily designed for unstructured data as
introduced in previous section, is being configured and managed using a different
portal interface. Compared to Persistent Storage, Object Storage does not require the
user to specify the amount of allocated storage in advance. In other words, as the
storage demand grows, the allocated capacity will simultaneously increase to
accommodate the requirements.
In order to use Object Storage, a Networked-Attached Storage (NAS) application is
required. This application has been installed on the virtual machines for each
organization. Figure 4.7, shows an overview of the created Object Storages:
Figure 4.7: Object Storage overview
43
Each Object Storage record has been given a unique username and password along
with an application key that would enable secure connectivity between data centre and
the virtual machines with installed NAS application.
Figure 4.8: Object Storage child account and authentication
Creating and configuring virtual machine instances for each organization
The following virtual servers were created as shown in Figure 4.6. The K Club and
translation.ie originally were offered both Microsoft Windows Server 2003, as well as,
Microsoft Windows Server 2008 virtual machines. O’Brien PR, however, has been
assigned a Microsoft Windows Server 2003 virtual machine. The Figures 4.5, showed
an overview of the created virtual machines and an example of virtual machine
technical specification is subsequently in Figure 4.9:
Figure 4.9: Detailed example of virtual machine specification
Installation on Cloud NAS involved a set of steps. As mentioned earlier, Cloud NAS
was necessary in order to make use of the Object Storage. A specific version of
Microsoft Framework (V3.5, SP3) was required to be installed on the created virtual
machines. Additionally, modification to the participating organizations firewall
configurations was required in order to allow for inbound and outbound traffic
associated with the cloud NAS application. Authentication was implemented through
the storage dedicated name and unique username/password, as well as, the application
key (Figure 4.10). This would enable the Object Storage to be envisioned as an
integrated part of the virtual machines. In this way, the allocated Object Storage
appears as a mapped disk drive on the virtual machine.
44
Figure 4.10: Client NAS authentication
Linking storage to the private network of the organizations
Two primary methods were used to link between the participating organizations’
network, and the pilot cloud platform. The first method is based on mapping the
provided storage using the static IP address assigned to the storage. This can be
implemented in various ways in different operation systems. For example, in its most
simple form, in windows operating systems, the “Map Network Drive” wizard can be
used to provide a step-by-step guide to the procedure of mapping a disk drive to/from
the network. To provide security measures, access to the drives are only possible if
correct authentication is provided.
Another access method is to use Remote Desktop Protocol (RDP) to the static IP
address assigned to the virtual machine instance. To provide security measures for this
model, RDP access on is only granted to specific users group defined in advanced.
However, according to Microsoft, RDP is vulnerable to “Man In The Middle Attack”
(MITM). In order to overcome to this potential security risk, administrators can enable
transport layer encryption (Microsoft, 2008).
To avoid data loss, an online backup solution (KeepITSafe) is also installed to back-up
the pilot system from IBM’s Datacentre. The back-up task is scheduled on a daily
basis and will be used in the event of data being accidently removed or if data
corrupted.
4.4 Design of survey study
In addition to the implementation of a cloud-based pilot system, as detailed in earlier
sub-sections, a survey study was additionally conducted aiming to discover perceived
advantages and disadvantages of adopting cloud computing amongst SMEs. As
described in chapter 2, the role of a CEO/CIO is of extreme importance when it comes
to adoption of new IT technology. For this very same reason, the survey study has
been divided to two categories, one specifically designed for C-level management
45
positions and another for IT users. Considering IT users, the main driver for them to
support migration to another IT technology is how that particular technology can make
their work routine easier, compared to the current technology that is being used. In
other words, simplicity and performance are much more important to the users than
other aspects, such as, security. For example, considering the top 10 security threats for
SMEs’ data (Pinzon, 2008), it can be seen that at least 5 directly relates to employees.
4 . 4 . 1 I T u s er s s u r v ey
Subsequently, the survey first tries to establish aspects of ICT technology that
employees are not currently comfortable with or have the most difficulties. These
aspects are: hardware, software (including operation systems, as well as, particular
applications), networking, telecommunication infrastructure and finally IT support
procedure. The findings of this part of the survey can indicate which cloud migration
strategy can be best supported by employees. For example, considering an SME where
the IT hardware facilities (including servers) are outdated and often causing services to
be unavailable, employees may be eager to see the CEO to move some of the
troublesome servers to cloud. Another aspect is the level of IT support. Depending on
their size and line of business, many SMEs may not be able to afford to have a full
time in-house IT support staff. In case of outsourcing the IT support, the level of
responsiveness, as well as, the pricing scheme amongst external contracting companies
varies significantly from one to another. If employees are not happy with the IT
support they received, they might be more open to moving to a cloud solution where
the cloud provider is responsible for service and maintenance of wide range of
software and hardware issues. In other words, employees prefer to be engaged with
their own work rather than being additionally responsible for troubleshooting their IT
problems. The questionnaire is designed in an interactive way so that the participant is
asked questions based on the answer to the previous questions. For example, if a user
indicates no experience in having hardware difficulties, no future question associated
with hardware will be presented.
The following are two sample questions that address the mentioned aspects:
46
Sample Question1:
Please choose the categories of which you have been experiencing difficulties the
most.
Please choose any applicable option
( ) Hardware; (e.g printer, mouse, keyboard, desktop etc.)
( ) Software; (e.g Microsoft office etc.)
( ) Network; (e.g shared drives, Internet connection etc.)
( ) Telecommunication; (e.g telephone, Mobile etc.)
Sample Question2:
What is your overall satisfaction with your IT help desk service?
( ) Very Satisfied
( ) Satisfied
( ) Neutral
( ) Dissatisfied
( ) Very Dissatisfied
The rest of the questions in this survey, try to find out the readiness of the
organizations infrastructure, as well as, the employees, for cloud adoption.
Subsequently, the survey includes questions, such as, familiarity of users with the
technologies, such as, virtualization or VPN that can be a part of a cloud solution. The
following are two examples of survey questions evaluating the awareness and
readiness of the organization for the cloud migration:
Sample Question1:
How familiar would you say you are with what cloud computing is?*
( ) Very familiar
( ) Somewhat familiar
( ) Not very familiar
Sample Question2:
Have you heard of the following cloud computing technologies?*
Yes No
IaaS (infrastructure services, such as, storage, network
capacity - compute)
( ) ( )
PaaS (Complete operating system and software package
available via cloud services- Develop)
( ) ( )
SaaS (Individual software packages) ( ) ( )
4 . 4 . 2 C I O s / C E O s s u r ve y
The second survey addresses the CEOs/CIOs of SME organizations. The survey first
establishes the size and business line of the correspondents, as well as, the role of the
participants in the organizations. Additionally, since the term ‘SME’ covers a diverse
range of organizations, it is important to categorise the correspondent organizations to
47
smaller sub-groups based on their number of employees. (See the following sample
questions)
Sample Question 1:
Please identify your primary role in connection in the organization or in
connection with ICT.
( ) CEO/Director/Chairman
( ) CIO/CTO/IT Manger
( ) IT Administrator/IT Staff
( ) other:
( ) Prefer not to say
Sample Question 2:
What is the size of your organization?
( ) 1-9 Employees
( ) 10-50 Employees
( ) 50-100 Employees
( ) 100-200 Employees
Subsequently, a set of questions are designed to investigate the perceived advantages
and barriers of cloud computing from the SMEs’ CEOs/CIOs perspective, as well as,
their view of difficulty of cloud migration should they decide to do so.
Sample Question 3:
Which of the following would you perceive as main barrier to you using Cloud
services?
( ) Security
( ) Service availability concern
( ) Initial CAPEX investment
( ) Provider lock-in
( ) Loss of control
Sample Question 4:
Do you expect migration to cloud services to be an easy adaption for your
organizations' users?
( ) Yes
( ) no
Finally, a set of questions were designed to reflect the option of the SME’s
CEOs/CIOs regarding the preferred cloud service type, as well as, other
considerations, such as, cost, security, and access flexibility.
48
Sample Question 5:
Do you envision security and privacy as concern for deployment of cloud
computing in your organization?
( ) Yes
( ) No
Sample Question 6:
Should you decide to deploy cloud computing services for your organization,
which of the following service types you would consider?
( ) IaaS
( ) PaaS
( ) SaaS
4.5 Conclusion
This chapter described the design and implementation of the cloud-based pilot system
used through this study. The reason behind choosing those three SMEs for the test-bed
evaluation as well the choice of the cloud provider and services have been described.
Subsequently, the steps involved in the implementation of the pilot system were
explained. Moreover, the chapter described the objectives of the two survey studies
conducted as part of this research. Furthermore, sample questions that aimed to
investigate some of those research objectives were presented. Chapter 5 will
investigate the outcome of the pilot project. It will additionally present and discuss the
results of the two survey studies described earlier.
49
5 EVALUATION, RESULTS AND DISCUSSION
5.1 Introduction
Using different performance metrics and evaluation methods, this chapter overviews
the result of the pilot system, as well as, the survey study described in the previous
chapter. The chapter is divided to two major parts. The first section, focuses on
evaluation of the pilot system. Such an evaluation is achieved through both qualitative
and quantitative analysis. While the qualitative evaluation is based on organization’s
feedback, the quantitative benchmarking focuses on networks’ latency aspects.
The second section presents the result of the survey study, and is divided in two parts.
The first part overviews the result of the survey which was designed for ICT users,
whereas the second part, focuses on the survey that addresses CEOs and CIOs. The
results of both surveys are analysed and the key findings of the study are further
explained.
5.2 Pilot system Evaluation
This section provides an evolution of the experimental pilot system used through this
study. The evaluation is divided to two parts. The first part employs a qualitative
approach, for evaluation of the pilot and hence is based on feedback from users who
have been using the service during the evolution period. The second part uses a
qualitative approach, and captures some performance aspects of the pilot through two
demonstrations.
5 . 2 . 1 Q u a l i t a t i ve a n a l ys i s
Table 5.1, summarises the details of the pilot implementation for the three participating
organizations. As Table 5.1 illustrates, O’Brien PR had been evaluating more services
and used more storage capacity compared to the other two organizations. As will be
explained further, translation.ie was found to be the most conservative organization
and had been only using the cloud pilot system as a file server that was hosting minor
part of their in-house network attached storage. Finally, the evaluation has been going
for four weeks for The K Club with six participating users. It is also interesting to
compare the Internet bandwidth of the three organizations.
50
0 1 2 3 4 5 0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
File Size (MB)
C D
F
Translation.IE
O'BrienPR
The K Club
Table 5.1: Pilot implementation and SME’s participation details
Organization Duration of Evaluation
Services Experimented Number of Users
Maximum Downlink Internet
Bandwidth
Utilized Storage
Frequency of Use
O’Brien PR 3 weeks
Terminal Session Access Remote eMail access
File Server Backup Storage
Remote Application Access (Adobe Photoshop)
2 2 Mbps 86.45 GB Medium
TranslationIE 3 weeks File Server
2 8 Mbps 0.57 GB Low
The K Club 4 weeks File Server
Backup Storage 6 30 Mbps 27.32 GB Medium
Alternatively, Figure 5.,1 shows the Cumulative Distribution Function (CDF) of the
size of the files stored on the cloud servers from each participating organization. In
fact, the CDF can illustrate the ratio of the files that are less than a given size shown on
the x-axis. For example, for the case when the file size equals to 1MB, the Figure 5.1
shows the CDF value of 0.55 for O’Brien PR. This in fact means that 55% of the files
that are uploaded to the storage server by O’Brien PR are smaller than 1MB and so on.
Based on this description, it can be easily seen that O’Brien PR has evidently much
larger files followed by The K Club. This is mainly because many of the files migrated
by the O’Brien PR to the storage servers are multimedia and image files that are large
in nature. To generate the results shown in Figure 5.1, a modified version of the
MATLAB™ program presented in (StackOverflow, 2010) is used to collect
information on the size of all the files stored under a specific directory name. The CDF
has been then generated using a MATLAB™ built-in command (cdfplot) that is
available on the MATLAB™ statistics tool box.
Figure 5.1: Cumulative distribution function of file sizes for participating SMEs
51
Regarding the access and usage statistics, Figure 5.2, shows the number of login
events using the RDP protocol for the O’Brien PR. This organization was the only
participant that used RDP.
Figure 5.2: Number of logins to the pilot on a daily basis for O'Brien PR
From Figure 5.2, it can be seen that the number of login events has been varying
noticeably during the evaluation period. As for the file servers and backups, Figure
5.3, shows the aggregated uplink and downlink bandwidth utilised by each
organization. This refers for example to the case when remotely stored files are
accessed through locally mapped disk drives (e.g. when the user tries to open the files
etc).
Figure 5.3 : Utilized bandwidth on a daily basis for O'Brien PR
7/6 9/6 11/6 13/6 15/6 17/6 19/6 21/6 23/6 0
5
10
15
20
25
Date [6 -24 June]
N u m
b e r
o f
L o g in
E v e n ts
7/6 9/6 11/6 13/6 15/6 17/6 19/6 21/6 23/6 0
100
200
300
400
500
Date [ 6-24 June ]
U ti li z e d B
a n d w
id th
( M
B )
52
Alternatively, Figure 5.4 and Figure 5.,5 show the corresponding utilised bandwidth
for the translation.ie and The K Club.
Figure 5.4: Utilized bandwidth on a daily basis for translation.ie
Figure 5.5: Utilized Bandwidth on a daily basis for The K Club
The following parts present the evaluation of the pilot project for each individual
organization.
5/6 7/6 9/6 11/6 13/6 15/6 17/6 19/6 21/6 0
2
4
6
8
9
Date [4 -22 June]
U ti li z e d B
a n d w
id th
( M
B )
1/6 3/6 5/6 7/6 9/6 11/6 13/6 15/6 17/6 19/6 12/6 23/6 25/6 0
7
14
21
28
35
Date [ 31 May -26 June ]
U ti li z e d B
a n d w
id th
( M
B )
53
A. O’Brien PR
O’Brien PR found to be enthusiastic towards migrating to cloud. The organization
initially planned to move many of the existing services to cloud. This included storage
and backup systems, as well as, a wide range of applications, such as, Microsoft Office
family products, Multimedia and Image processing applications. Due to the nature of
organization’s business as a public relations service company, there is especial need
for multimedia and image processing applications. Some of the key services offered by
the organization are considered as 1) Press service centre 2) DVD production 3) Print
and design management 4) Media relations 5) Photography organization and direction
6) Media training and 7) News press broadcast monitoring.
Because of the small size of the organization, the existing IT infrastructure found to be
small and simple consisting of a number of outdated computationally limited
workstations, a number of printers, scanners and a single file server. Additionally, the
company uses a broadband service with a relatively modest available bandwidth (peak
download bandwidth 2Mbps).
Observation and evaluation
Some of the examining services can be named as File Server, Back-up, Remote eMail
Access, Terminal Session Login, and use of some report application installed on the
pilot, such as, Adobe Suite (mainly Adobe Photoshop), Table 5.1, gives an overview
on the number of users involved and duration of pilot study.
The first issue noticed by the organization during the migration process was the time
needed for uploading the existing file server to the cloud storage. For transfer of nearly
87 GB of data (which is considered to be very small data chunk for a file server) to
cloud, the required estimated time was in orders of days (nearly 4 days). Moreover, the
copy process was repeatedly interrupted due to the network problems (See Figure 5.6).
54
Figure 5.6: Example of a network error
Regarding the applications installed, and specifically those used for multimedia
processing purposes (e.g. Adobe Photoshop), the users found it difficult to use those
applications in an interactive way through RDP. Based on the users’ feedback, latency
was a disturbing factor which was making a task that required graphical interactions
with this the software application to be inefficient and time consuming. Figure 5.7,
shows an example where the application environment, as well as, the operating system
environment (being Microsoft Windows 2008 Server herein) is not fully loaded.
(a)
55
(b) Figure 5.7: Examples of pilots’ disturbing factors form users’ perspective
Subsequently, the connection to the server was occasionally disconnected which
required the connection to be re-established before the user can resume the on-going
task (See Figure 5.8).
Figure 5.8: Examples of users’ experience on pilot system dis-connectivity
However, because of the improved resources (CPU and RAM) available on the server
platforms compared to the in-house workstations, the users found it faster when they
executed batch editing and automated processing of media files using the server
resources. This for example included resizing a large number of pictures or automatic
change of the pictures’ contrast etc.
56
Regarding the storage and file server used, there were indications of dissatisfaction
related to the latency of accessing the files that are stored on the remote cloud storage.
“The delay in loading files and especially multimedia files was a bit annoying, we are
a small company with few employees and we need to deliver our projects quick and
swift. We want to get the jobs done as fast as possible” said one of the employees who
have been using the file server.
After nearly 3 weeks of using the pilot system, a short interview with the Aileen
O’Brien, the company CEO and owner, was arranged on June 23 rd
2012. The final
conclusion was that the company sees some potential advantages in using the cloud-
based pilot system but they do not see the necessity to migrate to cloud any time soon
although they may use some form of cloud-based system for backup purposes. “….I
am not an expert in the computer domain, but based on my observations and the
feedback from my employees, I think we do not benefit from cloud much.” “...It might
be good for our backup system, but I think we can use more user friendly and cheaper
applications like Dropbox for that” said Aileen. When mentioned, she was unaware
of the fact “Dropbox” is classified as a cloud-based storage. This might indicate that
organization owners with less IT proficiency or background may see cloud computing
as a very complex system while they might have already used it in some form or other
and not be aware of it.
When asking about the organization’s plan in future to deploy cloud-based-system, the
answer was rather negative. “…honestly, if I want to invest on our IT, I prefer to
upgrade one or two of our own computers. I think that will be less expensive and more
useful for us.” said Aileen O’Brien with a laughter.
B. Translation.ie
Translation.ie was chosen for this study because of the fact that they have been dealing
with rather sensitive information. This includes translation of official documents and
evidence that used by Garda or offering translation services to individuals which also
needs a high level of confidentiality. According to the company founder and CEO,
“Privacy and confidentiality” is considered as a core part of the organization business.
“…we can assure all of clients that their confidentiality is firmly protected by our
adherence to the conditions set out by our ISO 9001:2008 and EN 15038:2006
certifications.” Said Crystal Li.
57
For this reason, the company agreed to move only a small part of their less sensitive
data to the storage server. After few weeks, it was observed that the data has been
accessed on few occasions. Since the storage disks were locally mapped to the users’
workstations, one of the users did not even notice that those files have been relocated
onto a remote storage server. Because of limited use of pilot cloud system, not much
performance feedback could be taken from the users. In general, the users didn’t feel
much difference when they were using those specific files compared to the past and
hence did not raise any performance issue. This was additionally due to the fact that
compared to the O’Brien PR, both the available bandwidth was much improved and
the files were smaller (See Figure 5.1) making it the transfer faster.
Additionally, an interview with Ronan Power (IT Administrator) was arranged on June
22 nd
2012. The following represents some key and related parts of the interview:
Mr. Power, do you have any concern regarding migrating some or all of our files and
data on to a cloud-based server?
“Yes, confidentiality is our main concern and we are concerned about unauthorised
access to our client’s data. That can lead to us losing our whole business.”
How do you assure confidentiality of you data in your existing IT infrastructure?
“I do not know what you mean exactly by ‘IT infrastructure’ but we have a pretty
much simple model in our company. Most of our translators and linguists are trusted
contractors and they sign our Confidentiality Agreement and our Code of Conduct
before accepting any job assignment with us. We normally ask them to remove all
related files from their computers once they finish an assignment. We then keep two
copies of all assignments on our external hard drives (pointing to one of the external
hard drives) and try to keep them safe.”
Talking from an IT administrator point of view, keeping all data in the office may not
be the best option. Imagine an incident, such as, fire; do you have a mechanism for
data recovery in that case?
“Then I should keep one of the disks at home (with laughter)!”
58
Yes, and that needs to be updated on a very regular basis if you want your latest
version of data. Don’t you think this will be time consuming with hassles?
“But at least it is free. Isn’t it?!”
What about access flexibility? Imagine you or some of your translators want to have
access to some of the data when they are away.
“Well, that can be useful but may be not a priority for us.”
Based on the conversations, it can be seen that the owner prefers simple and manual
based solutions, such as, using hard drives from backups and is very concerned about
the confidentiality of their client data.
C. The K Club
For The K club, a large portion of day-to-day working files were moved to cloud.
However, because the pilot system has not been fully tested, migration of more
sensitive information to cloud, such as, financial spread sheets and clients’ information
was avoided. Given a recent fire incident occurred in the organization server room, the
enterprise fully appreciate the importance of off-site back-up systems. Additionally,
the broadband speed used by The K Club was much higher than that of the O’Brien PR
company. This resulted in improved quality of experience (as result of decreased
latency) by the users who has tested the pilot system. Moreover, the users were not
involved in any of the configuration tasks needed for using the cloud-based storage.
The files that that were migrated to the remote server were expected to be used
regularly by six different employees. Before the migration, the files were located on a
local file server within the organization. The users were informed of the changes and
were ask to monitor any change of performance they would experience with the
system. After nearly four weeks of evaluation, the users were asked a series of
questions on June 26 th
2012. The questions and the users’ feedback are presented as
follows.
Did you notice any change in terms of access latency when accessing the files? Please
rate this on scale of 1 to 5 where 1 represents no latency difference and 5 represents
very significant latency (response shown in Figure 5.9).
59
Figure 5.9: The K Club’s user experience response – Latency experience
Did you ever have any issue accessing the mapped drive on your workstation (e.g.
receiving error message when trying to browse that drive)? (Response shown in Figure
5.10)
Figure 5.10: The K Club’s user experience response – Network mapping
Did you try to see if you could access to the files from a location outside the office
(e.g. home)? (Response shown in Figure 5.11)
Figure 5.11: The K Club's user experience response – Out of office access
1 2 3 4 5 0
1
2
3
4
5
User Rating
N o o
f U
s e rs
No Yes 0
1
2
3
4
5
6
7
N o o
f U
s e rs
No Yes 0
1
2
3
4
5
6
7
N o o
f U
s e rs
60
Further question revealed that the user who has been trying to connect to the file server
from another computer outside the company had not been able to connect to the file
server. It is suspected that this is due to the mal-configuration of the settings despite a
relatively easy to follow pictorial guide was provided to all users.
The above observations reveal that improved broadband quality could increase the
users’ quality of experience as four out of the six users indicated no or small latency
experienced when accessing files. Obviously, this very much depends on other factors,
such as, the type and size of the files they were trying to access, as well as, the
perceived definition of individual users from the latency as for example two users
experiencing the same level of access delay may rate the latency differently. This is
especially true given the very limited number of users feedback considered in this
study. Subsequently, in order to have a better understanding of the performance
metrics, it is vital to benchmark some of these metrics quantitatively.
5 . 2 . 2 Q u a n t i t a t i v e a n a l ys i s
Many aspects of the pilot system cannot be benchmarked to a quantitative means, this
for example includes concerns regarding the overall security and privacy. However,
there are performance aspects that can be benchmarked quantitatively. More
specifically speaking, the file I/O operation latency and computational latencies are the
main aspects that can be measured and compared against traditional IT infrastructure
frameworks. Subsequently, two demos illustrating extreme scenarios are presented.
The demos are in line with the requirements of one O’Brien PR company discussed
above.
The first demo, considers a scenario where an application requires consistent and
frequent access to graphical files and requires those files to be modified. In this
scenario, migration of the storage server (consisting of the working files) to cloud may
decrease the performance of such applications as a result of increased delay in
accessing (read and write) of the working files. Hence, migration of the storage
directory to cloud for such type of applications may not be very advantageous (unless
other solutions, such as, temporary caching of files in a local server are provisioned).
The second demo, concentrates on showing the advantages of a cloud system as a
result of availability of increased computational resources compared to a traditional IT
61
infrastructure in that individual workstations have limited computational power. For
this demonstration an application which requires heavy computation tasks is
considered and it is shown that the availability of computational resources in cloud as a
result of resource aggregation may improve the system performance in terms of
computational latency.
Demo1: Access latency for traditional and cloud-based applications
Unlike many large organizations who can afford to have very high capacity Internet
connections (e.g. T-CXR links), SMEs are normally having limited network
connections through normal DSL technologies.
As it can be seen, DSL is a dominant Internet access technology amongst all SMEs.
Despite of successful penetration of DSL technology during the last decay and
improvements in terms of available bandwidth and speed, the network capacity,
may still be considered as a limiting factor for SMEs when it comes to cloud
computing
62
that is heavily dependent on network connections. Based on the report from ComReg,
more than half of the SMEs are unaware of their downlink bandwidth and very few of
them had estimation of their available uplink bandwidth (ComReg, 2009).
In order to demonstrate the effect of access latency in a cloud solution, a demonstration
has been designed. This demonstration is in line with the requirement of the O’Brien
PR that required frequent access and modification of multimedia and image files.
For this propose, a Matlab™ based application has been developed in that the
application reads a series of image files (in JPEG format), modifies the brightness level
of the file and saves the resulted images as new files. The files are written to the
same location where the original image files are read from. To compare the latency
associated to such standard I/O operations, three scenarios were considered. In the first
scenario, the files are located on the local hard disk. Subsequently, the second case is,
when the files are stored on a fileserver that is part of the local network of the
organization (in-house fileserver). Finally, the last scenario is, when files are read
(written) from (to) remote cloud-based storage server.
Figure 5.14 shows the result of the broadband speed test performed on at the O’Brien
PR office where the effective downlink speed is around 1.45 Mbps and the
Uplink capacity is around 0.38 Mbps. As it can be seen, considering today’s DSL
services, the available network bandwidth is rather limited.
Figure 5.14: O’Brien PR Internet Bandwidth
63
To further evaluate the Internet connection available, the throughput, data and service
quality of O’Brien PR’s connection also was studied. The download/upload speed
were fully assessed and investigated to understand the fluctuation over time. The test
has declared rapid changing of speed indicating some data impact issues (Figure 5.15).
While, applications such as eMail will not be effected, multimedia services will be
degraded as a result of such variations in the Internet service.
Download speed Upload speed Figure 5.15: Download/upload speed
Furthermore, since data is transferred across the Internet in “blocks”, there will be a
pause between blocks being transferred known as TCP Pause. The TCP pause between
receiving subsequent blocks of data was examined for O’Brien PR’s connection to
understand possible network delays as a result of congestion or other network
problems. Figure 5.16, illustrates more steady behaviour in the downlink compared to
the uplink where significant delay samples can be observed.
TCP download Pause TCP upload Pause Figure 5.16: Download/upload TCP Pause
Alternatively, to evaluate the impact of the available broadband bandwidth, The K
Club was considered in another set of experiments. Figure 5.17 shows the
corresponding speed test result from The K Club office where the downlink speed is
64
increased to 24.41 Mbps and the uplink has been over six times faster compared to the
O’Brien PR network.
Figure 5.17: The K Club Internet Bandwidth
Variations of the available bandwidth over the time both for uplink and downlink, as
well as, the TCP Pause measurements for The K Club and translation.ie are illustrated
in the appendices.
Moreover, Figure 5.18 shows 4 of the sample files used in this demonstration.
Figure 5.18: Original files for latency demonstration
Figure 5.19 shows an example of the output file corresponding to the Figure 5.18
where the overall brightness of the image is decreased by 50%.
65
Figure 5.19: Output files as a result of latency demonstration
Figure 5.20, shows the size of all seven files (in Mega Bytes) used in this demo.
Figure 5.20: Size of sample files used for latency demonstration
During the experiments, the local file server was connected to the local machine
through a high speed switch (100mbps). Subsequently, Figure 5.21 shows the overall
latency of the application’s procedure (reading the files, modification and storing the
result file) in that more than 98% of the overall latency is due to file I/O operations,
The figure in fact, shows the overall latency for the three scenarios explained above.
As expected, the latency is by far the highest when the files are stored on a remote
storage server in cloud. Evidently, the minimum delay corresponds to the case when
files are accessed from the local disk. Obviously, the cloud latency significantly
1 2 3 4 5 6 7 0
2
4
6
8
10
12
J o b S
iz e (
M B
y te
s )
Job Index
66
depends on the available bandwidth and hence, there is a considerable difference
between the two cloud-based scenarios.
Figure 5.21: Latency comparison of Local Disk, Local File Server & Cloud
Demo 2: Computation latency for traditional and cloud-based applications
Resource aggregation is a fundamental element of cloud computing systems that
provides a cloud subscriber with significantly increased computational power (same
applies for storage). A central selling point for penetration of cloud to the SMEs’
market could be that it enables SMEs to have access to all these computing resources
without necessarily having to own them. The pay per use pricing model is a key
herein. For example, from time-to-time an organization may need to perform some
very CPU intensive tasks. The frequency of such requirement may be not too often
which makes it difficult to justify the cost of buying high power workstations for that
purpose. Alternatively, the organization can use cloud resources when needed and only
pays for the actual use of resources.
In-line with the scenario described by the O’Brien PR company, this demonstration
considers a case when the application requires performing a set of predefined CPU
intensive operations on the image files. For this purpose again a MATLAB™ program
has been developed in that image files are read and an edge detection operation is
performed on each of the input images. More specifically, MATLAB reads an image
1 2 3 4 5 6 7 0
20
40
60
80
100
120
140
160
Job Index
L a ta
n c y (
s )
Remote Cloud, DL=24.41Mbps UL=2.48Mbps
Remote Cloud, DL=1.45Mbps UL=0.38Mbps
Local File Server
Local Disk
67
as a Matrix and returns a binary matrix of the same size in that 1’s represents the areas
where an edge is detected and anywhere else is zero. The edge detection operation has
been considered as an example because, the functionality can be realised using
MATLAB built-in functions and that edge detection is considered as one of the most
CPU intensive operations (Kong et al., 2010). Figure 5.22 shows an example of an
input and output file considered in this demonstration.
Figure 5.22: Computation Latency demonstration input and output file
Additionally, to evaluate the required overall computation time as a function of the
available allocated CPU power, an emulator application is used. BES – Battle Encoder
Shirase (v1.5.2) is an application used in this study to limit the allocated CPU (as a
percentage of total CPU resources) to each of the running processes. This way, a
traditional, outdated workstation can be considered equivalent to the case when the
total allocated CPU of a high power machine is limited. Figure 5.23 shows two
snapshots of the BES – Battle Encoder Shirase application.
(a)
68
(b) Figure 5.23: Battle Encoder Shirase CPU Configuration a)identifying target process b)applying CPU
constraints
Figure 5.24 shows the computation time required to complete the task for the eight
image files used in the previous experiment. The experiments were performed on a
workstation with an Intel Xeon 3.06 GHz processor.
Figure 5.24: Linear interpolation and actual measured latency for computation task
In addition to the measured latency, a prediction based on a linear adjustment of
computation time versus the ratio of the allocated CPU was calculated and plotted. In
the predicted model, the total computation time is assumed to be increased by factor of
x when the allocated CPU is decreased by the same factor. As an example, considering
100% allocation as a reference point, this means that 10% CPU allocation should
results in increase of the total latency by factor 10 (240s). However, the measured
10 20 30 40 50 60 70 80 90 100 0
50
100
150
200
250
300
350
Allocated CPU (%)
T o ta
l C
o m
p u ta
ti o n T
im e (
s )
Measured Latancy
Linear Interpolation
69
results show that the latency is always above the simple linear projection. Overall, as
expected, the computation time improves with the increase of available CPU power.
While using the cloud-based system with improved available resources may decrease
the required computation time, the result of the computation may need to be
transferred back to the local machines which could results in some latency.
5 . 2 . 3 K ey o b s e r v a t i o n s o f t h e p i l o t e va l u a t i o n
Based on the evaluations of the pilot project the following observations were made:
Internet access is a key element that impacts the users’ quality of experience
when cloud-based services are used.
Interactive services, such as, RDP-based applications (as opposed to storage for
example) are evidently more sensitive to the latency.
Concerns around the data confidentiality and security can become a barrier for
cloud adoption for some SMEs.
The importance of off-site back-up systems and benefits of flexible remote
access to data may not be acknowledged by some SMEs.
Users may not show much interest to connect to the cloud storage from other
locations than the organization.
Obviously the above observation may not be considered as facts and conclusions since
very limited number of organization and users were considered in this study.
Subsequently, they should be treated as observations that can attract the attention of
the cloud providers to some potential concerns and some potential attitude from SMEs’
users/managers towards cloud computing technologies.
5.3 Analysis of survey study
This section overviews the result of the two survey studies described earlier. The first
sub-section (5.3.1), focuses on the survey designed for IT users whiles the second sub-
section (5.3.2), analyses the result of survey which has been completed by SMEs
CEO/CIO.
70
5 . 3 . 1 E va l u a t i o n o f I T u s e r s s u r v ey
As explained in the previous chapter, this survey aims to investigate two major aspects
namely the satisfaction of the IT users from the existing IT infrastructure and secondly
the readiness of the organization and employees for migrating to cloud.
Starting with the users’ satisfaction from the existing IT infrastructure, a series of
questions were presented to assess the structure of the organization and the IT
department and to identify areas where users have mostly encountered difficulties and
problems. The questions addressing these aspects are listed below together with the
correspondents’ feedback. Note that in some questions more than one answer choice is
allowed and hence sum of individual answers’ contributions may exceed 100%.
Figure 5.25 : IT Department availability for SMEs Total responses: 129
It is interesting to observe that most participating SMEs have their own IT
department/staff in-house (Figure 5.25). This can be changed when cloud migration
happens. Depending of the type service offered by the cloud provider, the SME might
decide to downsize or shut down its in-house IT department. This is because the cloud
provider will need to provide the support together with the offered service.
.
Figure 5.26: IT support team availability
Total responses: 128
71
Considering that majority of SME’s employees use phone and email to contact to their
IT support team, migration to cloud may not impose significant behavioural changes
for the users, as such migration may lead to outsourcing of the IT department in many
SMEs, where the cloud provider normally take care of offered services and
infrastructure (Figure 5.26).
Figure 5.27: Users’ satisfaction level of IT support services
Total responses: 110
Figure 5.27, shows that the IT users are in general positive about the level of support
they receive from their current IT support team. Given that cloud providers are not in-
house, they need to be highly responsive to the users’ requests to maintain this level of
users’ satisfaction. On the other hand, because of improved management over both
hardware and software aspects of the offered services from the cloud provider, it is
expected that users encounter difficulties less frequent.
Figure 5.28: Frequently occurring IT issue categories
Total responses: 114
As Figure 5.28 illustrates, the fact that majority of IT issues in SMEs are related to the
network and connectivity may not be seen as a promising sign for cloud migration in
72
that, depending on the cloud solution, the availability of the services or even the
infrastructure is heavily dependent on the network. The rate of IT issues around
hardware and software also seems to be nearly equal.
Figure 5.29: Frequently occurring hardware issues
Total responses: 46
While printing sound to be the main source of hardware issues for many employees
(Figure 5.29), it is not possible to move any of such services to cloud as the user
would need to collect their print outs locally. Additionally, due to the increased
flexibility provided by laptops and portable devices, SMEs may want to keep such
devices in use. Even with a cloud-based infrastructure, users will still need terminals to
connect to cloud and hence cloud is suspected not to be a replacement for existing
laptops used by employees. That is especially true considering those staff who need to
travel for business purposes on a frequent basis. Desktop computers, on the other hand,
do not offer any flexibility in terms of access to data/services, and hence, can be
replaced by cloud infrastructure provided that employees are equipped with terminals
(e.g. portable devices or laptops) that can access the cloud environment.
Figure 5.30: Frequently occurring software issues
Total responses: 48
73
Considering that majority of software issues are related to the operating systems
(Figure 5.30), PaaS services from the cloud providers may be appreciated since the
provider will be in charge of maintaining the operating systems. This includes assuring
regular updates, conflict resolution, applying security patches and protecting the
operating system against viruses and other threats. Considering business applications
on the other hand, decision regarding the cloud migration will be heavily dependent on
availability and support of the applications on the cloud environment. While there is
increasing efforts towards making most application compatible with the cloud model,
there are still many applications that may not be executed through cloud.
The second set of questions, tries to evaluate the level of familiarity of users with
cloud computing and its related technologies. The later can be achieved by establishing
to what extend those related technologies are currently being used within the
organization.
Figure 5.31: Users' familiarity with cloud computing
Total responses: 71
Figure 5.31, shows that while most IT users have heard about cloud computing
technology, due to the vague definition of cloud by the industry and the fact that the
technology is still in its early stages of development, most users consider themselves
not to be fully aware of all aspects of the technology. There is also nearly 17% of users
who have not come across cloud computing. However, it should be noted that the
transition to cloud may remain transparent for many users, especially those with
limited IT background (i.e. to use cloud services, a user may not need to be aware of
the underlying technology details).
74
Figure 5.32: Security & privacy concerns of cloud deployment (IT users)
Total responses: 64
Figure 5.32 illustrates results of an important question, which reveals that security and
privacy is the most important and critical concern of SMEs when cloud computing is
considered. Compared to the end users, the intensity around security issues is even
worse from the perspective of IT administrators or CIOs who are responsible personnel
for protecting the company.
Figure 5. 33: User familiarity with VPN
Total responses: 65
As per results in Figure 5.33, a vast majority of users are familiar with the VPN
technology. Since security is a major concern for cloud computing technology, the use
of VPN might be a necessity after migration to cloud. This is especially true
considering the maturity of the VPN technology and its flexibility including support
for IPsec and SSL for example. As an example of commercial VPN solutions
developed for cloud environment, mention can be made to VPN-Cubed® that enables
customer control in a cloud, across multiple clouds, and between private infrastructure
and the clouds.
75
Figure 5.34: VPN Security concerns
Total responses: 69
Results in Figure 5.34, show that while VPN technology can address the security
concerns of most users; however, there are still more than 20% of users who are
sceptical about the VPN technology and its robustness against security treats.
The following points summarise the key findings of the first survey study designed to
be answered by SME’s IT users:
IT users are in general satisfied with their current level of IT support they
receive. Subsequently, the cloud provides need to sustain users’ satisfaction.
Majority of IT issues are related to connectivity and the organization’s network.
This can be a concern when moving to cloud, in that availability of services and
infrastructure may be heavily dependent on the network.
Most software issues are related to the operation systems. This can make the use
of PaaS services more likely in that a cloud provider assures regular upgrades of
the operation system and resolving any conflict problem in the operation system.
Majority of the SME employees have indicated that they are to some extend
familiar with cloud computing technology.
Security and privacy issues are identified as deal-breaking concerns for cloud
migration by 63% of the users.
Vast majority of users, are familiar with the VPN technologies and most of them
are not concerned about security issues when VPN is in use.
76
5 . 3 . 2 E va l u a t i o n o f C I O s / C E O s s u r ve y
This survey aims to investigate different aspects of cloud migration and the factors
affecting the decision of SME organizations’ policy makers in regards to cloud
computing technologies. The first set of questions aims to identify type of participating
organizations and position of the survey correspondents.
Figure 5.35: Size of participating organizations
Total responses: 101
As it can be observed in Figure 5.35, the majority of participating SMEs have less than
50 employees while there exists also representatives of medium sized organizations in
the survey.
Figure 5.36: Participants’ role in connection with the organization
Total responses: 109
Fortunately, the participants consists of both CEO/Owners, which may not necessarily
be familiar with IT technologies, as well as, CIOs and IT administrators who are more
77
informed regarding the evolution of the IT industry and more specifically, cloud
technology that is the focus of this study (Figure 5.36). The next set of questions tries
to recognise the view of the organizations’ policy makers regarding the cloud
technology. More specifically concerns and perceived benefits of cloud computing are
investigated.
Figure 5. 37: Internet Bandwidth
Total responses: 64
As it is shown in Figure 5.37, most SME organizations are using connections with a
bandwidth lower than 16Mbps. Depending on the type of service being used, the
bandwidth may not be sufficient to support cloud services.
Figure 5.38: The main driver for cloud migration
Total responses: 86
As illustrated by Figure 5.38, cost reduction is the main driver for cloud migration for
SMEs. This has an important implication for cloud providers in terms of their pricing
strategies. In other words, in order to have a successful penetration in the SMEs market
78
(that is considerably large), they need to be able to come up with a pricing strategy
that, regardless of any performance consideration, is competitive with the cost of
traditional IT solutions. This should also leave some margin for the SMEs should they
require upgrading their Internet bandwidth and network infrastructure that are essential
for cloud services. The competition between the cloud service providers who are
growing in numbers can also lead to lower-priced offerings. Additionally, SMEs are
expecting to have an improved performance when migrating to cloud. This could be to
some extent due to the fact that if cloud providers offer an up-to-date service (both
hardware and software aspects) and that the integration and conflict issues in the
existing IT systems is resolved. According to MIT technology review, if there is a
reliability, or scalability issue, cloud providers can patch their platforms quickly and
address the problem (Cloud Expo, 2009). “This is because the issue will be in their
core business; so theoretically, they should be on their toes more so that an enterprise
IT team with only 5%-10% of the corporate budget tied to IT.” Outage incidents and
major reported issues with big cloud players may impact the perceived advantages of a
cloud-based system in terms of performance improvement. Thirdly, SMEs’ managers
and IT policy makers expect to have a simplified management system after migrating
to cloud. This to a large extend, is due to the fact that most cloud offerings are of
managed services type. Managed services refer to outsourcing daily IT management
responsibilities as a strategic method for improving the overall organizational
performance. Finally, access flexibility has been indicated as the least important driver.
This could be because, even without cloud, there is a mechanism that can prove access
flexibility.
Figure 5.39: Security & privacy concerns of cloud deployment (C-level perspective)
79
Total responses: 92
This simple question and results shown in Figure 5.39, conveys a clear message that
vast majority of decision makers are to some extends concerned about security and
privacy of cloud systems. According to Talbot, the security concerns around cloud
computing can be classified to two broad categories: security issues related to the
cloud providers and those that are associated to the customers (Talbot, 2010).
Figure 5.40: Main barriers of cloud computing identified by C-level/IT
Total responses: 58
Evaluation of Figure 5.40, confirms that security is the main concern when it comes to
cloud systems. Additionally, fear of service outage seems to be high amongst SMEs.
The service outage can be as a result of network issues and connectivity problems to
cloud, or it could be caused by cloud providers. However, the cloud outage is tidily
linked to the security issues, as the main source of cloud outage is unauthorised access
(Reed, 2010). Moreover, SMEs also seem to be to some extend worried about long-
term contracts with cloud providers and extreme dependency on them and being
unable to use another provider without significant switching costs.
The next series of questions aim to investigate the future plan of SME organizations in
terms of deploying cloud services. This includes the type of service that they are likely
to deploy, deployment time line and other considerations.
80
Figure 5.41: Hosting production/mission critical applications in cloud
Total responses: 75
One potential reason for the above observation in Figure 5.41, could be due to the
security concerns around the cloud systems. This assumption is in line with the
previous observation during the pilot system development in that the translation.ie was
found to be very conservative regarding storing sensitive information on cloud.
Figure 5.42: Most famous cloud service acquired to deploy
Total responses: 85
Figure 5.42, reveals that IaaS is by far the most popular cloud service amongst SMEs
followed by SaaS.
81
Figure 5.43: Cloud computing investment plan
Total responses: 59
Despite all positive sensations around cloud computing, the results in Figure 5.43, may
not indicate that SMEs are hugely enthusiastic about immediate deployment of cloud
solutions. However, it should be noted that the industry is still in its early development
ages and that SMEs might not be the frontiers of new technology deployments.
Moreover, the interest towards cloud technology is evidently increasing which can be a
positive sign.
Figure 5. 44: C-level/IT expectations from cloud migration
Total responses: 83
Figure 5.44, shows that decision makers with SMEs, are not too concern about the
transition phase, when migrating from the existing platform to a cloud-based one. This
includes the support and comfort of users after cloud-based solutions are deployed. In
fact, considering users’ perspectives, such transitions may be completely transparent to
82
the users. Other aspects include the consistency of IT system after migration and
temporary service disruptions.
The following points summarize the key findings of the second survey study designed
to be answered by SME’s CIO/CEOS:
Cost reduction is the main driver for deploying cloud-based system for most
SMEs.
Vast majority of SMEs have security concerns regarding the cloud deployment.
Subsequently, privacy and security issues are seen as the main barriers towards
cloud deployment.
74% of SMEs will consider IaaS as service, if they want to deploy a cloud-based
solution for their organization.
Only 10% of SMEs in the study were planning to deploy some sort of cloud-
based technology for their organization within the next 12 months.
62% of the CEOs expect cloud migration to be a smooth transition for the users
of the company.
5.4 Conclusion
The first section focuses on evaluation of the pilot system. Such evaluation is achieved
through both qualitative and quantitative analysis. While the qualitative evaluation is
based on organization’s feedback, the quantitative benchmarking focuses on networks
and computational latency aspects. The main findings of the study are that firstly,
access latency has shown to be a disturbing factor significantly affecting the users’
quality of experience. Two important factors in that regard are the bandwidth available
to the organizations and volume and size of the files in case of storage services.
Additionally, considering an SME with sensitive client information, it was revealed
that the organization were not very keen to use cloud storage due to the data
confidentiality concerns. Moreover, feedbacks from the organization owners show that
cost is a major factor when those SMEs consider cloud computing as an option. Cloud
services were perceived to be more expensive than the traditional IT system.
The second section presents the result of the survey study, and is divided in two parts.
The first part overviews the result of the survey, which was designed for ICT users,
83
whereas the second part, focuses on the survey that addresses CEOs and CIOs. The
key findings of both surveys were presented and discussed during the chapter. In
summary, despite that majority of users indicated some level of familiarity with cloud
computing, security and privacy concerns still seem to be existing around cloud
services. Origination owners and CIOs/CEOs however looked at the cloud deployment
from a cost reduction perspective but yet only a small percentage of SMEs are
planning to deploy cloud service in 2012/2013. The next chapter will conclude the
thesis.
84
6 CONCLUSION AND REMARKS
6.1 Introduction
Despite all the excitement and fuss around cloud computing technologies in recent
years, it is believed that the industry has been to some extent lagging behind when it
comes to the deployment of cloud technologies and migration from legacy IT
infrastructure. This research aims to investigate the influential factors that impact
adoption of cloud-based technologies by Small and Medium Enterprises (SMEs). To
achieve the objectives of the research, a cloud-based pilot system was designed and
implemented and had been evaluated by three SMEs from different sectors. In
addition, two survey studies were conducted to complement the research findings and
to overcome some limitation faced for the pilot evaluation. This chapter summarises
the key findings of the study, overviews the limitation of the research and describes the
plan for the future works.
6.2 Research Ov ervie w
This study investigated adoption of cloud computing technologies within Small and
Medium Enterprises (SMEs). The first chapter overviewed the scope of the work, the
research questions and the methodologies used in this study. Chapter 2, provided a
background to cloud computing including a formal definition of the term, type of
services available and deployment models of cloud computing. Subsequently, previous
literature studies addressing factors that impact adoption of IT solutions in general and
cloud computing in particular are reviewed in Chapter 3. Based on this review, the
main factors commonly mentioned by different authors are listed as the following:
1. Owner/CEO
2. Resources
3. Employees
4. Competitors
5. Suppliers and Customers
6. Government
85
The first three factors are intra-organization factors and the rest are external.
Considering internal factors, since many of the SMEs are based on the owner/manager
model, the role of the owner is of especial importance when IT strategies are decided
for the organization. In that sense, the openness and innovativeness of the CEO, as
well as, his/her IT proficiency define the attitude towards new IT technologies.
Additionally, SMEs are normally having more restricted financial constraints
compared to larger organizations. This could results in more conservative attitudes
towards deployment of new technologies. Moreover, the support of the employees for
adoption of a new technology can make the technology transfer to become smoother.
Regarding cloud computing, the main potential deployment driver reported by other
studies is the cost reduction. Additionally, concerns around security and privacy have
been found to be a major main barrier.
Chapter 4, described the design and implementation of the pilot system used through
this study. The reason behind choosing those three SMEs for pilot system evaluation,
as well as, the choice of the cloud provider and services were also described.
Subsequently, steps involved in implementation of the pilot project were explained.
Moreover, the chapter defined the objectives of the two survey studies conducted as
part of this research and presented sample questions that aimed to investigate some of
those research objectives.
Finally, Chapter 5, presented the evaluation of the cloud-based pilot system and
discussed the results of the conducted surveys. As for the pilot evaluation, the first
company, O’Brien PR, has indicated that despite of some interruptions that were
mostly due to the network errors, the pilot system has been functional. The employees
of O’Brien PR have been using the pilot system both for the storage and back-up, as
well as, running their applications using RDP. As for the back-up, the initial upload of
the data to the cloud storage has been found to be a very lengthy process (in orders of
days) that is additionally interrupted by occasional errors when files could not be
copied. Fortunately, this procedure (initial upload of bulk data) may not need to be
repeated on a daily basis, and hence can be ignored if the incremental daily data update
is smooth (as it was the case for O’Brien PR). However, the large network latency had
made it almost impossible for users to execute their applications through RDP. The
86
application in use, was the Adobe Photoshop that is used for processing of image files.
According to the users, large network latency had made it very difficult to use the
application in an interactive way, while due to the availability of the increased CPU
and memory resources on the cloud platform (compared to local workstations), batch
file processing was faster in cloud. A short chat with Aileen O’Brien the owner of the
organization revealed that they were not very keen to use cloud for running their
applications but would consider that as an option for storage, whilst prefer Dropbox for
that matter due to its simplicity, previous experience of use and more importantly cost.
Regarding the second company, translation.ie, the organization agreed to move only a
very small part of their data to the cloud storage. This was mainly due to the concerns
of the company about the security and confidentiality of their clients’ data. After three
weeks of evaluation, a feedback from users was collected and it was revealed that users
did not notice any difference in terms of the access latency, or availability of data. This
is expected to be associated to the increased availability of the network bandwidth,
compared to O’Brien PR case and the smaller size of the files that have been used. In
addition, a short interview with the CEO was arranged and it was discovered data
privacy is a major concern for the company, and that the organization had the tendency
to keep the legacy IT infrastructure as it has been “functional and sufficient” according
to CEO Crystal Li.
As for The K Club, investigation and users’ feedback show that four out of six users,
experience no or little latency when using the cloud storage. This is expected to be due
to the high network bandwidth available to the organization. Additionally, none of the
users had experienced any difficulty in terms of accessing the cloud storage disks that
were mapped on the local machines. Whilst the cloud-based storage of the pilot system
enabled users to access the file from different locations, only one user had indicated
that she has been trying to use the feature and access the files from home. Although
this observation may not be conclusive due to the short duration of the evaluation
phase, and limited number of users considered, it might suggest that, global access
may not be the most important benefits that users might expect from a cloud-based
storage.
87
Based on O’Brien PR’s experience, using the cloud-based storage of the pilot system,
two demonstrations were designed and implemented. In the first demo, a Matlab™
application is developed that reads image files, modifies the brightness level of the
images, and write them to the same directory. Subsequently, three scenarios were
considered. In the first scenario, the files are located on the local disk of the
workstation that runs the Matlab™ application. The second scenario refers to, when
files are on the local (in-house) file server, and finally third scenario is when they are
accessed from the remote cloud-based storage. The application captures the overall
latency starting from when it starts to read the files, until, when the modified files are
fully stored. As for the cloud-based storage scenario, results from The K Club cloud
storage is compared against that of the O’Brien PR. There is significant variation of the
captured latency when comparing the two cases. This is mainly due to the difference in
the available Internet bandwidth (both upload and download directions) of the two
organizations. The second demo emphasises on the computational latency and it shows
how the increased CPU availability in cloud can results in improved computational
latency. For this task again a Matlab™ application is developed that performs a CPU
intensive task (edge-detection) on images.
In general, based on the pilot system evaluation, it was found that, the access
bandwidth plays a very critical role in users’ quality of experience when cloud services
are deployed. Considering SMEs, where checking emails is considered as the most
important reason for using Internet, the Internet services need to be upgraded before a
major migration to cloud is considered as an option (Poon and Strom, 1997).
Subsequently, the results of the two surveys were presented and discussed in Chapter
5. The first survey was designed to be answered by organizations’ employees. The
survey results reveal that connectivity and network problems are the main issues that
users are dealing with. Given that most cloud services are dependent on the network;
this may not be a very optimistic indication. Additionally, 63% of employees had
indicated that security and privacy issues are deal-breaking concerns for cloud
migration. On the bright side though, vast majority of users have indicated that they
are to some level, familiar with cloud computing and have used VPN-based solution in
past.
88
The second survey that is responded by CEOs and CIOs of organizations, shows that
cost reduction will be the main objective should they consider migrating to cloud.
Additionally, vast majority of the respondents mentioned security as a major concern
when it comes to cloud solutions. Moreover, considering different cloud services, it
seems that IaaS is by far the most popular cloud service that SMEs are after. The
outlook for immediate transition to cloud services was shown not to be very optimistic
with only 10 of survey participants mentioned plan for cloud migration within the next
12 months.
6.3 Contributions to the Body of K nowledge
The following are the key findings of this research study that investigates different
aspects of cloud computing amongst SMEs.
Security and privacy issues are identified as deal-breaking concerns for cloud
migration by 63% of the users.
Cost reduction is the main driver for deploying cloud-based system for most
SMEs.
Internet access is a key element that impacts the users’ quality of experience
when cloud-based services are used.
74% of SMEs will consider IaaS, if they want to deploy a cloud-based solution
for their organization.
The importance of off-site back-up systems and benefits of flexible remote
access to data may not be acknowledged by some SMEs.
Majority of IT issues, are related to connectivity and the organization’s network.
This can be a concern when moving to cloud in that availability of services and
infrastructure may be heavily dependent on the network.
62% of the CEOs expect cloud migration to be a smooth transition for the users
of the company.
Only 10% of SMEs in the study were planning to deploy some sort of cloud-
based technology for their organization within the next 12 month.
89
The implications of the research findings for SME owners/managers can be noted as
the following:
SME owners need to better understand cloud services and perhaps the underlying
mechanisms to some extent and to have realistic expectations from different
cloud services. They also need to have a clear understanding of their
organization requirements before making any decision on particular cloud
offerings.
SMEs should pay enough attention to their existing network infrastructure
including, the internal network and their Internet links in terms of reliability and
available capacity, when deciding on a cloud offering. Upgrade of the
bandwidth/infrastructure, may be required before deploying a cloud solution.
SME owners need to investigate the security measures implemented by the cloud
providers before moving their sensitive data to cloud.
Subsequently, the implications of these findings for service providers can be listed as
the following:
Connectivity is an important issue that directly impact the quality of experience
when cloud services are considered. Strategic partnership of the Internet and
cloud service provides can result in mutual benefits for both parties.
Alternatively, Internet service providers are faced with a great opportunity to
enter the cloud domain.
Service providers need to quickly adapt the latest solutions that can assure SMEs
regarding their data security and confidentiality. Such solutions can be originated
from third party organizations, such as, Cloud Security Alliance (CSA) or
standardisation bodies.
The overall cost of ownership for cloud services targeting SMEs should be
maintained low.
Service providers need to increase the awareness and perceived advantages of
different cloud services amongst SMEs. This should particularly be focused on
services that currently are less known or unpopular.
90
6.4 Lim itation of the research
The limitations of this study are to a large extent related to the methodologies that are
used. Regarding the implementation and evaluation of the pilot project, one should
note that it may not be possible to implement the pilot system for more than three SME
organizations. This was mainly due to the cost constrains of implementation and the
fact that, not many organizations are keen to migrate some of their day-to-day IT
functionalities to a pilot cloud solution that can impose the risk of service disruption,
or compromise of the security and confidentiality of their data. Subsequently, although
maximum effort is given to select three SMEs from different sectors with different size
and IT requirements, still, the feedback coming from those organizations may not well
represent the SME’s opinions about the cloud experience. Additionally, the type cloud
services under evaluation, as well as, the choice of the cloud provider could impact the
experience of the users. Finally, not all aspects of a cloud-based solution could be
measured, either qualitatively or quantitatively.
For these reasons, and to be able to have larger and more diverse feedback from SMEs,
two survey studies were conducted. Fortunately, the number of respondents was in an
acceptable range (around 80 and 130 response for each survey). However, most
respondents appear to be from European countries, and therefore, the results may not
represent a world-wide view on cloud computing technology.
6.5 Future Work and Research
The following can be considered as potential research works that can extend this
research:
Detailed analysis of the cost aspects (both OPEX and CAPEX) and comparison
of the cloud offering and existing technologies for SMEs.
Investigation and evaluation of other cloud services that has not been considered
in this study (e.g. PaaS).
Comparison of different cloud deployment models for SMEs (e.g. Hybrid versus
Private) and to investigate the advantages and disadvantages of each model.
Detailed investigation, to find out, to what extend the current cloud offerings are
exposed to security threats when comparing against conventional technologies.
91
7 BIBLIOGRAPHY
Adamov, A., and Erguvan, M.,2009, The truth about cloud computing as new
paradigm in IT, International Conference on Application of Information and
Communication Technologies, pp. 1—3.
Agarwal, R., and Prasad, J., 2000, A Field Study of The Adoption of Software Process
Innovations by Information Systems Professionals, IEEE Transactions on
Engineering Management, Vol 47, No 3, pp 295-308.
Alam, S., 2009, ICT Adoption in Small and Medium Enterprises: an Empirical
Evidence of Service Sectors in Malaysia, International Journal of Business &
Management, Vol.4, No. 2.
Ambrust, M., Fox, A., Griffith, R., Joseph, A., Katz, r., Konwinski, A., Lee, G.,
Patterson, D., Rabkin, A., Stoica, I. and Zaharia, M., 2009, Above the Clouds:
A Berkeley View of Cloud Computing, UC Berkeley RAD Systems Labs.
Anon, 2008, Let it rise, A special report on corporate IT, The Economist, Vol. 389,
No.8603.
Armbrust M., et al., 2010, A View of Cloud Computing, Comm. ACM, vol. 53, no. 4,
pp. 50-58.
Arnold, J., 2010, Why Businesses Need to Think Differently About Cloud
Communications, Web: http://www.focus.com/briefs/why-businesses-need-
think-differently-aboutcloud, last visited on 13, April 2012.
Attewell, P., 1992, Technology Diffusion and Organizational Learning: The Case of
Business Computing, Organization Science, Vol 3, No 1, pp 1-19.
Bardin, J., Callas, J., Chaput, S., Fusco, P., Gilbert, F., 2009, Security Guidance for
Critical Areas of Focus in Cloud Computing v2.1, Cloud Security Alliance.
Barham, P., Dragovic B., Fraser, K., Hand, S., Harris, T., Ho, A., Neugebauer, R.,
Pratt, I. and Warfield, A., 2003, Xen and the art of virtualization, 19 th
Symp. on
Operating Systems Principles, p 164–177, Bolton Landing, USA.
Bose, R., and Sugumaran, V., 2006, Challenges for Deploying Web Services-Based E-
Business Systems in SMEs, International Journal of E-Business Research, 2(1),
1-18.
92
Buyya, R. & Venugopal, S. 2009. Market oriented computing and global Grids: An
introduction, in Market Oriented Grid and Utility Computing. John Wiley &
Sons. ISBN: 9780470455432.
Buyya, R. Broberg, J and Goscinski, A., 2011, Cloud Computing Principles and
Paradigms. John Wiley & Sons. ISBN 9780470887998.
Catlett, C. 2002, The philosophy of TeraGrid: Building an open, extensible, distributed
TeraScale facility, Proceedings of 2nd IEEE/ACM International Symposium on
Cluster Computing and the Grid, Germany.
Chau, P., 1995, Factor Used in The Selection of Packaged Software in Small
Businesses: Views of Owners and Managers, Information & Management, Vol
29, No.2, pp 71-78.
Cloud Expo, 2009, MIT Technology Review covers cloud security.
ComReg, MBIMS, 2009, SME & Corporate ICT Research, Commission for
communication regulation, 41108987 JMM/PM/SB, By Millward Brown IMS,
Page33.
Cragg, P., and King, M., 1993, Small-firm Computing: Motivators and Inhibitors, MIS
Quarterly, Vol 17, No 1, pp 47-59.
Creeger, M., 2009, CTO Roundtable: Cloud Computing, Communications of the
ACM, Vol.52, No. 8, pp. 50—56.
Davis, F., 1989, Perceived Usefulness, Perceived Ease of Use and User Acceptance of
Information Technology, MIS Quarterly, Vol 13, No 3, pp 319-340.
Dian, H., Ying, L. and Zhi, H., 2011, Replicate Distribution Method of Minimum Cost
in Cloud Storage for Internet of Things, Network Computing and Information
Security (NCIS), IEE International Conference.
Donelaicio, K., 2012, Cloud Computing Based Accounting for Small to Medium Sized
Business, Inzinerine Ekonomika Engineering Economics, 23(1), 14-21, ISSN
1392–2785.
Drew, S., 2003, Strategic Uses of Ecommerce by SMEs in The East of England,
European Management Journal, Vol 21, No 1, pp 79-88.
Duffy, S., 2010, Tourism And Hospitality Research In Ireland Conference, Shannon
College Of Hotel Management.
Durkee, D., 2010, Why Cloud Computing Will Never Be Free, ACM Queue, Vol. 8,
No.4, pp. 1—10.
93
Durkee, D., 2010, Why Cloud Computing Will Never Be Free, Commun. ACM, vol.
53, pp. 62–69.
Dutta, S., and Evrard, P., 1999, Information Technology and Organisation within
European Small Enterprises, European Management Journal, Vol 17, No 3, pp
239-251.
Earl, M.,1989, Management Strategies for Information Technology. Prentice Hall,
New York.
Etro, F.,2009, The Economic Impact of Cloud Computing on Business Creation,
Employment.
European Commission, 2001, The 2001 ENSR Survey on SMEs: The Seventh
Observatory of European SMEs, Brussels.
European Parliment Cloud Comuitng Study, 2012, Directorate General For Internal
Policies, Policy Department A: Economic And Scientific Policy, Cloud
Computing Study ,Ip/A/Imco/St/2011-18 , Pe 475.104.
Fink, D., 1998, Guidelines for The Successful Adoption of Information Technology in
Small and Medium Enterprises, International Journal of Information
Management, vol. 18, no. 4, pp. 243-253.
Fink, D., 2002, Building The Professional Services E-Practice, in Managing
Information Technology in Small Business: Challenges and Solutions, IDEA
Group Publishing, Hershey, pp 246-260.
Foster, I., Zhao, Y., Raicu, I., Lu, s., 2008, Cloud Computing and Grid Computing
360-Degree Compared, Grid Computing Environments, Workshop, DOI:
10.1109/GCE.2008.4738445.
Galliers, R., and Sutherland, A., 1999 ,Information Systems Management and Strategy
Formulation: Applying and Extending The Stages of Growth Concept in
Strategic Information Management , edition 2, Editors Galliers, R., Leidner, D.
and Baker, B., Butterworth-Heinemann, Oxford, pp 33-60.
Garsthagen, R., 2012, Oracle: Your Cloud, your choice,4th Cloud Comuting wolrd
Forum Computing, London.
Gartner, 2008, Gartner Says Cloud Computing Will Be As Influential As E-business,
Press release, Http://www.gartner.com/it/page.jsp?id=707508, last visited on
13, April 2012.
Geelan, J. 2008. Twenty one experts define cloud computing Virtualization, Electronic
Magazine, Available from: http://virtualization.sys-con.com/node/612375, last
visited on 25, February 2012.
94
Georgios I., Smithson, S., and Lybereas, T., 2001, Trends in information technology in
small businesses. Idea Group Publishing, UK, ISBN 9781930708044.
Harmon, R., Demirkan, H., Hefley, B. and Auseklies, N., 2009, Pricing Strategies for
Information Technology Services: A Value-Based Approach, 42nd Hawaii
International Conference on System Sciences, pp. 1—10.
Hollander, A., Cherrington, J., and Denna, E., 2000, Accounting, Information
Technology, and Business Solutions, Irwin/McGraw-Hill, Boston.
Huebscher, M., McCann, J., 2008 , A survey of autonomic computing—degrees,
models, and applications, ACM Computing Surveys , Volume 40 Issue 3,
Article No. 7 ,DOI:10.1145/1380584.1380585.
IDC , 2009, IDC Says Cloud Computing Is More Than Just Hype; Worldwide IT
Spending On Cloud Services Expected To Reach US$42 Billion By 2012,
PressRelease,Http://www.idc.com/AP/pressrelease.jsp?containerId=prSG21724
009, last visited on 23 June, 2012.
International Business Machines Corp, 2006, An architectural blueprint for autonomic
computing, White Paper Fourth Edition.
Jaakkola, H., 1996, Comparison and Analysis of Diffusion Models: Diffusion and
Adoption of Information Technology, Chapman & Hall, New York, pp 65-82.
Kazi, A., 2007., Small and Medium Business Enterprises and the use and adoption of
Information and Communication Technology: a study of Legal Issues and
Legal Perspectives. International Journal of Organisational Behaviour, 12(1),
144-160.
Keahey, K., Foster, I., Freeman, T. and Zhang, X., 2006, Virtual workspaces:
Achieving quality of service and quality of life in the grid, Scientific
Programming Journal, 13(4):265:275.
Khajeh-Hosseini, A., Greenwood, D. and Sommerville, I., 2010, Cloud Migration: A
Case Study of Migrating an Enterprise IT System to IaaS, ArXiv preprint,
submitted to 3rd IEEE International Conference on Cloud Computing.
Kong, J., Dimitrov, M., Yang, Y., Liyanage, J., Cao, L., Staples, J., Mantor, M., Zhou,
H., 2010, Accelerating MATLAB Image Processing Toolbox Functions on
GPUs.
Krauter, K., Buyya, R. and Maheswaran, M., 2002, SOFTWARE—PRACTICE AND
EXPERIENCE: A Taxonomy and Survey of Grid Resource Management
Systems for Distributed Computing, Softw. Pract. Exper., 32:135–164 ,DOI:
10.1002/spe.432.
95
Lanz, J., 2002, Worst Information Technology Practices in Small to Mid-Size
Organizations, The CPA Journal, Vol 72, No 4, pp 71-74.
Laudon, K., and Laudon, J., 2000, Management Information Systems: Organization
and Technology in The Networked Enterprise. Prentice Hall, Upper Saddle
River.
Lin, G., Zhu, J. and Dasmalchi, G., 2009, Cloud Computing: IT as a Service, IT
Professional, Vol. 11, No. 2, pp. 10—13.
Love, P., Irani, Z., Li, H., Cheng, E. and Tse, R., 2001, An Empirical Analysis of The
Barriers to Implementing E-Commerce in Small Medium Sized Construction
Contractors in The State of Victoria, Australia , Construction Innovation, Vol
1, No 1, pp 31-41.
Mehrtens, J., Cragg, P. and Mills, A., 2001, A Model of Internet Adoption by SMEs,
Information & Management, Vol 39, No 3, pp 165-176.
Mell, P. and Grance, T., 2009b, The NIST Definition of Cloud Computing. Version
15, 10-7-09, http://csrc.nist.gov/groups/SNS/cloud-computing/cloud-def-
v15.doc, last visited on 15, May 2012.
Mell, P., and Grance, T., 2009a, Effectively and Securely Using the Cloud Computing
Paradigm, Ver.26, Http://csrc.nist.gov/groups/SNS/cloud-computing/cloud-
computing-v26.ppt, last visited on 15, May 2012.
Mell, P. and Grance, T., 2011. The NIST definition of Cloud Computing. National
Institute of Standards and Technology, Spec. Pub, 800-145.
Microsoft, 2008, Configuring Terminal Servers for Server Authentication to Prevent
“Man in the Middle Attacks".
Mladenow, A., Fuchs, E., Dohmen, P. and Strauss, C., 2012, Value Creation using
Clouds, Analysis of Value Drivers for Start-Ups and Small and Medium Sized
Enterprises in the Textile Industry, 26th International Conference on Advanced
Information Networking and Applications Workshops.
Murch, R., 2004, Autonomic Computing IBM Press, ISBN-13: 978-0131440258.
OfCom, 2010, Office of Comminications, Annual Report and Accounts 2010.
Papazoglou, M. and Heuvel, W., 2007, Service oriented architectures: Approaches,
technologies and research issues, The VLDB Journal, 16:389, 415.
Pinzon, S., 2008, WatchGaurd: Top 10 Threats to SME Data Security.
96
Poon, S., Strom, J., 1997, Small Businesses' Use of the Internet.
Qian, L., Luo, Z., Du, Y. and Guo, L., 2009, Cloud Computing: An Overview,
Proceedings of 1st International Conference on Cloud Computing Conference,
pp. 626—631.
Reed, J., 2010, Following Incidents into the Cloud, SANS Institute.
Rimal, B. and Choi, E., 2009, A Conceptual Approach for Taxonomical Spectrum of
Cloud Computing, Proceedings of the 4th International Conference on
Ubiquitous Information Technologies & Applications, pp. 1—6.
Rogers, E., 1983, Diffusion of Innovations, New York.
Ronchi S., 2003, The effects of the Internet adoption in customer-supplier
relationships, ISBN 90-77360-02-6.
Saleh, S., Wang, C., 1993, The management of innovation: Strategy, structure, and
organizational climate, IEEE Transactions on Engineering Management, vol.
40, no. 1, pp. 14–21.
Sarosa, S. and Zowghi, D., 2003, A Strategy for Adopting Information Technology for
SMEs, Electronic Journal of Information Systems Evaluation, 6(2), pp 165-
176.
Sarosa, S., Zowghi, D., 2003, Strategy for Adopting Information Technology for
SMEs: Experience in Adopting Email within an Indonesian Furniture Company
, Electronic Journal of Information, 10th European Conference on Information
Technology Evaluation, Spain.
Schollhammer, H., Kuriloff, A., 1979, Entrepreneurship and Small Business
Management, ISBN : 978-0471762607.
Seyal, A., Rahim, M. and Rahman, M., 2000, An Empirical Investigation of Use of
Information technology Among Small and medium Business Organizations:
Bruneian Scenario, The Electronic Journal on Information Systems in
Developing Countries, EJISDC 2, 7, 1-17.
Sotomayor, B., Montero, R., Llorente, I. and Foster, I., 2009, Virtual infrastructure
management in private and hybrid clouds, IEEE Internet Computing,
13(5):14:22.
Srikantaiah, S., Kansal, A. and Zhao, A., 2008, Energy aware consolidation for cloud
computing, the 2008 conference on Power aware computing and systems, p
10-10, USENIX Association Berkeley, USA.
97
StackOverflow, 2010, How to get all files under a specific directory in MATLAB?,
http://stackoverflow.com/questions/2652630/how-to-get-all-files-under-a-specific-
directory-in-matlab, last visited on 17, June 2012.
Staten, J., Yates, S., Ryme, J., Gillett, F. and Nelson, L., 2009, Deliver Cloud Benefits
Inside Your Walls: Economic and Self Service Gains Are Within Reach,
Forrester Research Inc.
Staten, J., 2009, Cloud Is Defined, Now Stop the Cloudwashing, Systems Evaluation,
Volume 6, Number 2, pp.165-176.
Talbot, d., 2010, Security in the Ether,Web:
http://www.technologyreview.com/featured-story/416804/security-in-the-
ether/, last visited on 15, May 2012.
Taylor, S., Young, A. and Macaulay, J., 2010, Small Businesses Ride the Cloud: SMB
Cloud Watch—U.S. Survey Results, Cisco Internet Business Solutions Group
(IBSG).
Thong, J., Yap, C., 1995, CEO characteristics, organisational characteristics and
information technology adoption in small business, Omega, vol. 23, no. 4,
pp.429-442.
Thong, J., Yap, C., 1996, Information Technology Adoption by Small Business:
Diffusion and Adoption of Information Technology, Chapman & Hall, London,
pp. 160-175.
Tsai, C., Lin, U., 2011, Information Security of Cloud Computing for Enterprises,
Advances on Information Sciences and Service Sciences. Volume 3, Number 1.
Turban, E., MacLean, E. and Wetherbe, J., 2002, Information Technology for
Management: Transforming Business in The Digital Economy, John Wiley &
Sons, New York.
Uhlig, R., Neiger, G. and Rodgers, D., 2005, Intel virtualization technology, IEEE
Computer, vol. 38, no. 5, pp48:56.
Utomo, H. and Dodgson, M., 2001 ,Contributing Factors to The Diffusion of IT
Within Small and Medium-sized Firms in Indonesia, Journal of Global
Information Technology Management, Vol 4, No 2, pp 22-37.
Vaquero, L., Rodero-Merino, L., Caceres, J. and Lindner, M., 2009, A Break in the
Clouds. Towards a Cloud Definition, Computer Communications Review, Vol.
39, No.1, pp. 50—55.
98
Venkatesh, V., and Brown, S., 2001, A Longitudinal Investigation of Personla
Computers in Homes: Adoption Determinants and Emerging Challenges, MIS
Quarterly, Vol 25, No 1, pp 71-102
Vijeikis, J. and Makstutis, A., 2009, Small and Medium Sized Business
Competitiveness in Lithuania, Economy and Management: Issues and
Perspectives, 2(15), 328-338.
Vincent. L., Guynes, J., 1997, An assessment of the influence of organizational
characteristics on information technology adoption decision: a discriminative
approach, IEEE Transactions on Engineering Management, Vol44, Issue: 2
DOI: 10.1109/17.584923, PP146- 157.
Walczuch, R., Braven, G. and Lundgren, H., 2000, Internet Adoption Barriers for
Small Firms in The Netherland, European Management Journal, Vol 18, No 5,
pp 561-572.
Welsh, J. and White, J., 1981, A Small Business Is Not a Little Big Business, Harvard
University Business review, No. 81411, Vol 59, No. 4, pp 18-32.
Williams, J., 2009, NASA Nebula in Action: Cloud Computing Case Examples,
NASA Ames Research Center.
Wyld, D., 2009, The Utility Of Cloud Computing As A New Pricing, And
Consumption –Model For Information Technology, International Journal of
Database Management Systems, Vol. 1, No. 1, pp. 1—20.
Wyld, D., 2010, The Cloudy Future Of Government IT: Cloud Computing and the
Public Sector around the World, International Journal of Web & Semantic
Technology (IJWesT), Vol 1, Num 1.
Zhang, Q., Cheng, L. and Boutaba, R., 2010, Cloud computing: state-of-the-art and
research challenges, Journal of Internet Services and Applications, Vol. 1, No.
1, pp.7—18.
99
8 APPENDICES
8.1 Appendix list
Each Appendix item can be found on the accompanying Compact Disc, entitled,
“Cloud computing for SMEs – Challenges, Barriers & Outcomes”. The following list
outlines its contents:
Appendix
#
Title Document Format
A IT Users Survey Report PDF/Microsoft Word Format
B CIO/CEO Survey Report PDF/Microsoft Word Format
C
Internet Test result for O’Brien PR Graph
O'Brien PR Download Speed Test Results
O'Brien PR Upload Speed Test Results
PDF Format
Microsoft Excel Format
Microsoft Excel Format
D
Internet Test result for translation.ie Graph
Translation.ie Download Speed Test Results
Translation.ie Upload Speed Test Results
PDF Format
Microsoft Excel Format
Microsoft Excel Format
E
Internet Test result for The K Club
The K Club Download Speed Test Results
The K Club Upload Speed Test Results
PDF Format
Microsoft Excel Format
Microsoft Excel Format
F
Quantitative Evaluation Demonstration 1
Quantitative Evaluation Demonstration 2
Folder Containing:
Quantitative Experimental demonstration Media
Produced Media
BES Application
Matlab Files for Quantitative demonstration
G
Cloud-based Pilot Project files and samples
Folder containing:
Pilot Experiment vendor Contract
SME’s Pilot Certificates Logs
Installation applications
Object Storage Portal (HTML SAMPLE)
Persistent Storage portal (HTML SAMPLE)
A Sample invoice for part of the service
100
8.2 Appendix A
CIO/CEO Survey Report
Cloud Computing for SMEs - ( From an IT Manager or C-level
Position perspective)
What is the size of your organization?
Value Count Percent
1-9 Employees 30 29.70%
10-50 Employees 38 37.60%
50-100 Employees 10 9.90%
100-200 or more Eomployees 23 22.80%
Statistics Total Responses 101
Which industry sector does your organization belong to?
Value Count Percent
Agriculture 2 2%
Banking, Finance and Insurance 7 6.90%
Government 4 4%
Telecommunications 5 5%
IT Services 23 22.80%
Energy, Water and Mining 4 4%
Health and Education 8 7.90%
Leisure & Hospitality 4 4%
Construction 3 3%
Manufacturing and Engineering 9 8.90%
Property and Business Services 1 1%
Retail and Wholesale 11 10.90%
Transport, Storage and Distribution 3 3%
Services 8 7.90%
Other: 9 8.90%
Statistics Total Responses 101
Please identify your primary role in connection in the organization or in connection with ICT.
Value Count Percent
CEO / Director / Chairman 29 26.60%
CIO / CTO / IT Manger 33 30.30%
IT Administrator / IT Staff 36 33.00%
other: 3 2.80%
Prefer not to say 8 7.30%
Statistics Total Responses 109
101
Open-Text Response Breakdown for If other, please specify Count
HR 1
Admin 1
ordinary user 1
What type of internet access do you have?
Value Count Percent
Analog 0 0%
T-1 Lines 2 2.90%
DSL / ADSL / ADSL+2 37 54.40%
SDSL 2 2.90%
VDSL 1 1.50%
Cable 6 8.80%
ISDN / B-ISDN 7 10.30%
I Don't know 5 7.40%
OC3 4 5.90%
Satellite 1 1.50%
T-3 Lines 3 4.40%
What is your internet Bandwidth?
Value Count Percent
Up to 4Mbps 2 3.10%
Up to8 Mbps 27 42.20%
Up to 16Mbps 11 17.20%
Up to 30Mbps 5 7.80%
More than 30Mbps 14 21.90%
I Don't know 5 7.80%
Statistics Total Responses 64
What is your IT adaption philosophy towards technology?
Value Count Percent
Innovator 21 22.80%
Early Adopter 15 16.30%
Fast Follower 28 30.40%
Early Maturity 14 15.20%
Late Adopter 12 13%
I don't Know 2 2.20%
Statistics Total Responses 92
What is the main driver for your organisation to migrate into the cloud?
Value Count Percent
102
Reduced cost 41 47.70%
Simpler administration and maintenance 17 19.80%
Access flexibility 10 11.60%
Improved performance 18 20.90%
Statistics Total Responses 86
Do you plan to invest in cloud computing in short, medium or long term?
Value Count Percent
Short term (Within next 12 months) 6 10.20%
Medium term (within 1-2 years) 9 15.30%
Long term (more than 2 years) 25 42.40%
Uncertain 19 32.20%
Statistics Total Responses 59
Which of the following would you perceive as main barrier to you using Cloud services?
Value Count Percent
Security 20 34.50%
Service availability concern 15 25.90%
Initial CAPEX investment 6 10.30%
Provider lock-in 12 20.70%
Loss of control 5 8.60%
Statistics Total Responses 58
Has your organization ever experienced an Internet security breach that affected business operations?
Value Count Percent
Yes 12 18.80%
No 46 71.90%
I don't know 6 9.40%
Statistics Total Responses 64
Do you envision security and privacy as concern for deployment of cloud computing in your organisation?
Value Count Percent
Yes 84 91.30%
No 8 8.70%
I don't know 0 0%
Statistics Total Responses 92
103
Should you decide to deploy cloud computing services for your organisation, which of the following service types you would consider?
Value Count Percent
IaaS 63 74.10%
PaaS 5 5.90%
SaaS 17 20%
Statistics Total Responses 85
Would you host production / mission critical applications in a hosted Cloud solution?
Value Count Percent
Yes 10 13.30%
No 65 86.70%
Statistics Total Responses 75
If a decision is made for cloud services to be deployed in your organisation, would you use consultancy firms for the deployment?
Value Count Percent
Yes 7 8.90%
Depends on cost & consultancy service provided 39 49.40%
No 33 41.80%
Statistics Total Responses 79
Do you expect migration to cloud services to be an easy adaption for your organisations' users?
Value Count Percent
Yes 52 62.70%
No 31 37.30%
Statistics Total Responses 83
Which of the following statements is true in your opinion?
Value Count Percent
Cloud Computing is more expensive than Traditional Computing
42 48.80%
Traditional Computing is more expensive than Cloud Computing
44 51.20%
Statistics Total Responses 86
104
8.3 Appendix B
IT Users’ Survey Report
Cloud Computing for SMEs - (From an ICT user perspective)
Does your organization have a separate department responsible for IT?
Value Count Percent
Yes 88 68.2%
No 32 24.8%
I don't know 9 7%
Statistics Total Responses 129
When you are reporting an IT problem, do you know how you should Contact the Technical Support team?
Value Count Percent
Yes 106 82.2%
No 23 17.8%
Statistics Total Responses 129
How do you usually contact your IT Department or IT support to get help?
Value Count Percent
In person 33 25.8%
By phone 59 46.1%
By eMail 54 42.2%
By a Helpdesk 27 21.1%
Other 18 14.1%
Statistics Total Responses 128
Open-Text Response Breakdown for Other Count 3 I fix myself 1 I will call HR 1 There is also a Request Link where we log cases 1 We are a very small team 1 We resolve IT issues ourselves 1 Web portal 1 haven't gotten help 1 i dont know 1 i would email my manager 1 never had to worry about anything! 1 no IT 1
105
no IT man 1 there is no ITdepartment 1 we have an outsourced company 1 we manage this ourselves 1
To your best of knowledge, can you access any of the following systems or applications over the internet?
Yes
% #
Finance Systems /Human Resource & Payroll Systems 37.6% 44
Document & File Management Systems 46.2% 54
Assets Management Systems / Inventory Management Systems 26.5% 31
Decision Support Systems (DSS) 16.2% 19
CRM/SFA / ERP systems 26.7% 31
Buinsess Intelligence Systems / Collaboration Systems 29.9% 35
Disaster Recovery System 28.2% 33
Tele/video/audio conferencing / Communications 54.7% 64
Sales and marketing 53.4% 63
Access to remote database 51.3% 60
Project management (billing/ time/expense management) 36.8% 43
Manufacturing (scheduling/quality control) 23.9% 28
Average % 36.0%
In your opinion, how serious is the threat of outsiders or hackers gaining access to businesses' online information?
Value Count Percent
Major threat 46 39%
Moderate threat 46 39%
Minor threat 21 17.8%
No threat 5 4.2%
Statistics Total Responses 118
Please choose the categories of which you have been experiencing difficulties the most.
Value Count Percent
Hardware; (i.e: printer, mouse, keyboard , desktop etc...) 46 40.4%
Software; (i.e: Microsoft office etc...) 50 43.9%
Network; (i.e: shared drives, internet connection etc...) 68 59.6%
Telecommunication; (i.e: telephone , Mobile etc... ) 32 28.1%
Statistics Total Responses 114
106
What are the main hardware issues you are usually experiencing?
Value Count Percent
Desktop related 29 63%
Laptop and portable device related 23 50%
Printer related 31 67.4%
Cabling related 14 30.4%
Other 0 0%
Statistics Total Responses 46
Open-Text Response Breakdown for Other Count
Please indicate how satisfied you are with the IT assistance your are receiving in the following Hardware issues
Very satisfied
% #
Desktop related 33.3% 14
Laptop and portable device related 30.8% 12
Printer related 24.4% 10
Cabling related 25.7% 9
Which of the following software you usually have technical difficulties with or receive a lot of errors when using them?
Value Count Percent
Business related applications (i.e: Accounting software, purchasing or marketing software etc..) 28 58.3%
operating system (i.e: unable to logging on, receiving error when saving documents etc...) 32 66.7%
eMail and communication ( i.e: unable to send/received email, archiving issues, password issues) 27 56.3%
Other 4 8.3%
Statistics Total Responses 48
Open-Text Response Breakdown for Other Count 1 Microsoft office 1 TRC 1 ms 1
Please indicate how satisfied you are with the IT assistance your are receiving in the following Software issues
Very satisfied
107
% #
Business related applications 19.1% 9
operating system 10.9% 5
eMail & communication 19.6% 9
What are the main network issues you are usually experiencing?
Value Count Percent
Connecting to shared drives 43 63.2%
Surfing the web and accessing websites 30 44.1%
Poor quality of experience 43 63.2%
Other 4 5.9%
Statistics Total Responses 68
Open-Text Response Breakdown for Other Count Main system failing night runs or giving interface
errors 1 Systems Freezing 1 connecting to dDB 1 connecting to other computers 1
Please indicate how satisfied you are with the IT assistance your are receiving in the following Network issues
Very satisfied
% #
Connecting to shared drives 14.9% 10
Surfing the web and accessing websites 16.4% 11
Quality of experience 13.4% 9
What are the main Telecommunication issues you are usually experiencing?
Value Count Percent
Desk phone related 17 53.1%
Mobile & portable device related 16 50%
Wireless and Unified Communication related 24 75%
Video related 16 50%
Audio related 2 6.3%
Other 1 3.1%
Statistics Total Responses 32
Open-Text Response Breakdown for Other Count never had to be concerned ... 1
What is your overall satisfaction with the your IT help desk service?
108
Value Count Percent
Very Satisfied 23 20.9%
Satisfied 30 27.3%
Neutral 37 33.6%
Dissatisfied 15 13.6%
Very Dissatisfied 5 4.5%
Statistics Total Responses 110
Please indicate how satisfied you are with the IT assistance your are receiving in the following Telecommunication issues
Very satisfied
% #
Desk phone related 16.1% 5
Mobile & portable device related 23.3% 7
Wireless and Unified Communication related 30.0% 9
Video related 26.7% 8
Audio related 13.3% 4
Do you envision security and privacy concerns to be the main barrier for cloud deployment for your organisation?
Value Count Percent
Yes 40 62.5%
No 24 37.5%
Statistics Total Responses 64
Are you familiar with any of the following technical terms ?
Yes
% #
VPN ( Virtual Private Network ) 60.0% 66
Virtualization 45.5% 50
Cloud Computing 67.3% 74
Are you familiar with the Virtual Private Network ( VPN ) technology?
Value Count Percent
Yes 55 84.6%
No 10 15.4%
Statistics Total Responses 65
If a VPN is being used, which of the following business functions does your organization use a VPN connection for?
Value Count Percent
109
Using web-based office applications 25 39.7%
Tele/video/audio conferencing 18 28.6%
Access to remote database 26 41.3%
Communications 27 42.9%
Document management 29 46%
Sales and marketing 20 31.7%
Finance and accounting 13 20.6%
Supply chain management -inventory –purchasing 5 7.9%
Project management (billing/ time/expense management) 8 12.7%
Human resources/payroll 9 14.3%
Manufacturing (scheduling/quality control) 4 6.3%
Other 5 7.9%
eMail 32 50.8%
Statistics Total Responses 63
Open-Text Response Breakdown for Other Count 3 Not totally sure 1 none as far as I know 1
In your own opinion , what is the most important benefit of using VPN?
Value Count Percent
Less travel cost 9 14.1%
Less travel time 18 28.1%
Improved security 26 40.6%
Increase availability 11 17.2%
Statistics Total Responses 64
Do you still have security concerns when using VPN?
Value Count Percent
No 15 21.7%
No 54 78.3%
Statistics Total Responses 69
Have you heard of any of the following terms?
Yes
% #
Server virtualisation 92.0% 46
Desktop virtualisation 66.0% 33
Average % 79.0%
110
Have you felt any difference in terms of service , when connecting databases which you know they are running on virtual servers?
Value Count Percent
Service has been slower 6 12%
Service has been faster 9 18%
No change in Service 28 56%
Don't know if we use virtualisation 9 18%
Statistics Total Responses 50
How familiar would you say you are with what cloud computing is?
Value Count Percent
Very familiar 16 22.5%
Somewhat familiar 43 60.6%
Not familiar 12 16.9%
Statistics Total Responses 71
Have you heard of the following cloud computing technologies?
Yes
% #
IaaS (infrastructure services such as storage, network capacity - compute) 45.1% 32
PaaS (Complete operating system and software package available via cloud services- Develop) 29.6% 21
SaaS (Individual software packages) 64.8% 46
Which of the following cloud computing technologies do you think will be the most important one in the future?
Value Count Percent
IaaS 20 29.9%
PaaS 19 28.4%
SaaS 28 41.8%
Statistics Total Responses 67
Which of the following statements is true in your opinion?
Value Count Percent
Cloud Computing is more expensive than Traditional Computing 20 42.6%
Cloud Computing is less expensive compared with Traditional Computing 27 57.4%
Statistics Total Responses 47