Building a Knowledge Base for Guiding Users through the Cloud Life
Cycle
Claudio Giovanoli and Stella Gatziu Grivas
Institute for Information Systems, University of Applied Science, Northwestern, Basel, Switzerland
Keywords: Cloud Computing, Cloud Life Cycle, Knowledge Base, Resource Description Framework, Infoplace.
Abstract: Cloud Computing is still a hype to provide IT services as customers demand. Especially small and medium
size businesses can benefit of this new trend because of the flexibility Cloud Computing provides for the
consumption of its services. Nevertheless the step into the cloud must be carefully prepared and setup to
ensure a successful usage and an appropriate return on investment. Currently the lack of experience with
cloud projects and the uncertainty of its predestined application areas are detaining companies to use cloud
computing services. No best practices, rules and use cases are in place yet to help to convince potential
users to go their way to the cloud. With the CLiCk project the authors therefore propose a personalized web-
based platform for Small and Medium Enterprises to support them over an entire Cloud Life Cycle.
1 INTRODUCTION
Cloud Computing becomes more and more popular
also for small and medium enterprises (SME).
Besides all the well-known advantages of such
services, several entry thresholds, open questions
and concerns increase the initial and management
costs of a cloud project dramatically. During the last
2-3 years several studies have shown that the main
concerns are based on security, regulatory,
governance and availability issues (Armbrust et al.,
2010). From a SME perspective such issues are
leading to questions like: is my company Cloud
ready, how have other enterprises approached to the
cloud, how can the services be govern and
monitored or in which way can services be
terminated? Normally the appropriate know-how to
answer such questions is not available inside a
company and thus, to solve it external knowledge
(e.g. expertise by consultants) through an entire
cloud life cycle needs to be purchased for a high
price. They have to assess the risks, the benefits and
best fitting provider for the evaluation. Furthermore
they have to setup and monitor a migration plan and
governance and management of the Cloud services.
Finally, if a client wants to terminate the current
service, external expertise has to proof possibilities
of leaving and/ or moving the Cloud service in a
technical and judicial sense. Especially for SMEs
this is a hurdle which cannot be afforded due to the
lack of resources (most often caused by budget
constraints). This situation leads to the fact that such
entities are stuck between a rock and a hard place.
On the one hand they could benefit from cloud
services on the other hand the lack of resources does
not allow to evaluate and start its use.
Recent surveys are showing that companies are
aware about the advantages of cloud computing but
they are not ready to go in this direction caused by
the issues mentioned above. A survey conducted
from Cambridge Technology in Switzerland during
summer 2011 shows that 55% of all companies
going into the cloud need the support of external
consultants (Cambridge Technology, 2011). A
survey led from Easynet during spring 2011 shows
that in 42% of the companies the business is not
convinced of cloud computing and that 1/3 of all
CIOs has no framework to measure the ROI
(Easynet, 2011). A European wide study published
by Vanson Bourne (2012) explains that around 85%
of SME are doubtfully about the application and
infrastructure services provided through a cloud.
Additionally most named issues are security,
governance, data privacy and availability issues as
the main constraints.
Regarding Cloud Life Cycle G. Conway & E.
Curry (2012) have proposed a four phase lifecycle
with nine subsequent steps. These phases are very
similar to our Cloud Life Cycle. However we focus
131
Giovanoli C. and Gatziu Grivas S..
Building a Knowledge Base for Guiding Users through the Cloud Life Cycle.
DOI: 10.5220/0004375401310134
In Proceedings of the 3rd International Conference on Cloud Computing and Services Science (CLOSER-2013), pages 131-134
ISBN: 978-989-8565-52-5
Copyright
c
2013 SCITEPRESS (Science and Technology Publications, Lda.)
on the usability for SMEs and the application within
the CLiCk project with building and realizing the
appropriate artifacts. Instead of nine subsequent
steps our cloud life cycle is broken down into an
open number of different tools, so called artifacts
with the aim to support users to their cloud stories.
When it comes to cloud marketplaces and
brokering services the work of Buyya (2009) defines
as follows: Market Maker/Meta-broker is a part of
Cloud infrastructure that works on behalf of both
Cloud users and Cloud service providers. It
intermediates access to spread resources by
discovering appropriate cloud providers for a given
user application and attempts to optimally map
users’ jobs and requirements to published services. It
is a part of a global marketplace where service
providers and consumers join to find suitable match
for each other.
In this paper we introduce the CLiCk approach;
we give a brief overview of the architecture and the
several artifacts. In particular we depict the Cloud-
Life-Cycle with its four different stages covering
different scenarios that arise around cloud services
form a user perspective.
2 THE CLiCK APPROACH
The vision of CLiCk (Cloud Life Cycle) is the
provision of services und supportive information
which can be accessed on an appropriate platform
through the accordant enterprises.
During its first visit, the user has the ability to
create an account and made up a company profile on
the platform. At the beginning this profile contains
general information about the users’ company and it
will be enriched during the continuous usage of the
platform based on the user’s input. Based on this
profile, the user will be guided and supported
through its individual way through the Cloud Life
Cycle and it will receive personalized output from
the system. This personalization approach offers a
tailored support to the individual case and needs. For
example if a user is taking note that he would like to
use Infrastructure as a Service (IaaS), in the further
step of assessing its risks, the assessment will be
focusing on IaaS and not on e.g. Software as a
Service. One step further, the Provider and Service
Classification will only explain offerings for IaaS.
The services and information provided to the user
are collected in different repositories. These services
accompany the user over an entire cloud life cycle
and deliver a decision and management support on
common valid questions. Thus these repositories are
containing different types of artifacts.
We define the Cloud Life Cycle containing four
different stages which a user will go through while
thinking about cloud services and consuming it. The
Life Cycle model contains the phases, evaluation,
migration, operation and change. Each phase comes
with its own characteristics and prepares the user to
decide to move one step forward or not. The first
stage “Evaluation” brings the customer to different
areas which have to be analyzed. There the readiness
question needs to be answered, if the company is
ready to move to a cloud. Furthermore a Cloud
strategy has to be build and also important is to
choose the right partner for providing right the
service in the right time. After the cloud strategy is
set, the second part of the life cycle begins, the
Migration. Users have to prepare the migration into
a cloud. They setup a migration plan, including
fallback scenarios etc and fulfill the migration form
on premise the migration. While obtaining a Cloud
service it is recommended to manage the service.
Not in the sense of keeping the service running, but
to control the adherence of SLA’s, governance and
compliance issues. Finally if a user decides to stop
using the service or the change the service provider,
it comes to the “change” phase. Here the exit
strategy, part of a holistic Cloud strategy, has to be
applied. While the client decides to change the
provider, he has to do the contractual termination, to
ensure his data remains his data and to begin again
with the evaluation phase.
3 THE ARCHITECTURE
The current architecture level foresees several layers
of the CLiCk platform. At the top a web-based
interface enables the stakeholders to interact with the
system. The next layer, the “matching layer”
describes an Inference Engine. It derives based on
the information in the different repositories the
supportive knowledge. The following sub-chapters
are describing the current set of artifacts and the
corresponding repositories.
3.1 The Artifacts
To assess the users need and pain points, some
artifacts are collecting information about the current
as-is situation and the possible to-be plans. Based on
the assessed (e.g. readiness maturity) further
artifacts like next steps and advises, guidelines are
automatically offered to the user.
CLOSER2013-3rdInternationalConferenceonCloudComputingandServicesScience
132
Figure 1: CLiCk Big Picture.
Therefore the artifacts are based on assessments,
checklists, and guidelines. As for the different life
cycle stages the artifacts are varying. Two examples
are:
Risk Assessment
There are already several research works
investigating different risk areas of cloud computing.
In particular we rely on the seven points introduced
by Mather (2009): Infrastructure Security, Security
and Storage, Identity and Access Management,
Security Management in the cloud, Privacy, Audit
and Compliance, Cloud Service Provider.
A further scenario, described by Krutz & Vines
(2010), illustrates a very detailed listing of possible
risks like Confidentiality, Integrity, Availability,
Identification, Authentication, Accountability,
Authorization and Privacy.
In CLiCk we focus on a more simplified
framework which can be easily stored in the
knowledge base and going through the inference
process. We identify three main areas for all set of
cloud risks: ‘business risks’, legal risks’ and
technological risks’. While assessing the user’s risks
each of these parts will be appraised and feedback or
further recommendations will be displayed to the
user.
The Cloud Use Case Framework
Literature is showing that there exist different
methodologies for analyzing use cases. However,
these methodologies are either generic or specialized
but not for cloud computing services. Different
criteria which are considered important for
describing cloud services use cases will assess the
fitness of these methodologies. An evaluation of the
different methodologies has shown, that currently no
approach exists which covers all important aspects
of a cloud use case (Schwitter, 2012). Thus, for
providing a holistic use case and to deliver an
optimal support to the users, the authors have
decided to build up a use case framework from
scratch. To consider important aspects (technical and
business relevant) four different layers are identified
to describe and analyze a use case.
3.2 Data Bases and Repositories
Provider Data Base
The provider database will contain information on
the different service providers. It reflects a so called
provider landscape, where the suppliers are
categorized under certain aspects. Furthermore the
services offered by each provider are evaluated and
will be stored here too. The provider database is the
basement for the provider- and service classification
which will be shown to the user as a result of its
different assessments.
Knowledge Base (KB)
The central repository of CLiCk is the so called
knowledge base. It contains most of the different
introduced artifacts and represents the knowledge of
the tool. As the artifacts are interconnected, the
relations between the artifacts have to be considered.
Taking this into account the artifacts and its data
have to be stored in a structured way. Thus it is
foreseen that the artifacts and its output within this
KB are modeled with the Resource Description
Framework (RDF). The RDF approach has been
chosen for establishing more flexibility for further
developing and integrating new artifacts. It also
empowers the possibility of describing the
customer’s need and to match them with the
different opportunities. The following example gives
a short overview on the RDF based output of the
IaaS Readiness Assessment. It shows the question,
its description and a possible answer. Regarding the
chosen type of given answers, an appropriate first
feedback is given to the user.
Use Case Repository
The use case repository enables to store the collected
cloud use cases. It follows a developed framework
which defines different areas of interest within such
a use case. Following this scheme establishes also
that use cases can be compared on the different
topics like the service- and deployment model but
also on technical and management issues. The use
cases should be (i) a viable source for the user to see
how other have compete their cloud projects and (ii)
to support the user by identifying different
BuildingaKnowledgeBaseforGuidingUsersthroughtheCloudLifeCycle
133
workloads / process areas which are predestined to
run in a cloud.
3.3 Matching
Whereas the Use Case Repository can be handled
with a entity-relation model, the Provider DB and
especially the Knowledge Base need a more
sophisticated approach. It may be very often the
problem that the query to retrieve the appropriate
provider or piece of knowledge is too general
respective to specific resulting into a big result set
resp. empty result. The retrieval mechanisms need to
suggest possible specializations of the query resp.
possible relaxations to the query. Thus the
architecture has foreseen a matching system (i) that
has to cope with inaccurate matching results and (ii)
has to be adaptable for upcoming requirements. The
matching system will be able to determine
appropriate specializations of a given query. The
system will consider which further specializations
are (i) very common to the user in general as well as
(ii) for him specifically depending on a determined
user model and (iii) which promise a good reduction
of the huge result set.
In case of an empty set the matching system will
analyze which part of the query can be changed in
which way in order to retrieve at least one result. For
example, if someone is looking for a cloud provider
with a specific business service but can’t find it,
probably there is a provider with a more generic or
related business service, which can serve as a
substitute. The matching system will realize the
relaxation and suggest the provider with an explicit
explanation why this provider is retrieved.
4 CONCLUSIONS
AND OUTLOOK
The introduced CLiCk approach is a first attempt to
build up smarter Infoplaces. Its main intention is to
offer self-services to SMEs for assessing their cloud
needs and abilities. Whereas big companies have the
possibility to gain such know-how through
consultancy, Small and Medium Enterprises are
mostly being left on their own, due to budget
constraints.
Based on the knowledge base, other different
repositories and the user’s (companies) profile an
e.g. Prolog based inference engine combines the
given facts and derives personalized output to the
user to support him through the entire cloud life
cycle.
As the high level architecture of the platform is
now almost finished and first artifacts too, the
authors are now starting to cope with the detailed
concept. Afterwards a first prototype will be
implemented.
Thinking one step further the authors see an
opportunity, while including the provider- and
service landscape to shift the entire platform from a
pure information source to a cloud market broker
service. According to Buyya (2009) it reflects a part
of a global marketplace where service providers and
consumers join to find suitable match for each other.
It provides various services to its customers such as
resource discovery, accounting and pricing services.
In contrary to Buyya this marketplace focus on the
need and pain points of users interested in cloud
services. Once assessed the system evaluates the
most convenient services and service providers can
be through the system automatically derived and
suggested to the user.
REFERENCES
Vanson Bourne, (2012). Manage your changing IT needs,
Survey, Dell Business Solutions, Retrieved from:
http://www.dellsbsolutions.com/Main/BusinessLine/B
usinessLine.aspx?step=8&ItemID=17.
Easynet, (2011). Umfrage zum Cloud Computing,
Unternehmen vergessen Netzinfrastruktur, Survey.
Cambridge Technology Partners, (2011). Cloud
Computing in Switzerland, A Survey on the Adoption
of Cloud Computing by Swiss Companies, Zurich:
CTP-Consulting.
M. Armbrust et al. (2010), A View of Cloud Computing,
Comm. ACM, vol. 53, no. 4, pp. 50–58.
Conway, G., and Curry, E. (2012). “Managing Cloud
Computing: A Life Cycle Approach,” In 2nd
International Conference on Cloud Computing and
Services Science (CLOSER 2012): Porto, pp. 198–
207.
R. Buyya et al., (2009). Cloudbus Toolkit for Market-
Oriented Cloud Computing'Proc. 1st Int'l Conf. on
Cloud Computing, Beijing, 1-4 December 2009.
Krutz, R.L. & Vines, R.D., (2010). Cloud Security: A
Comprehensive Guide to Secure Cloud Computing,
Hoboken: Wiley.
Mather, T., Kumaraswamy, S. & Latif, S., (2009). Cloud
Security and Privacy: An Enterprise Perspective on
Risks and Compliance, O’Reilly Media, Retrieved
from:http://books.google.com/books?id=BHazecOuD
LYC&pgis=1.
D. Schwitter, (2013). A customer centric view and use
case analysis of the current cloud service situation in
Switzerland, Master Thesis, Olten: University of
Applied Sciences and Arts Northwestern Switzerland.
CLOSER2013-3rdInternationalConferenceonCloudComputingandServicesScience
134
