Adapting Service Development Life-cycle for Cloud
George Feuerlicht
1,2,3
and Hong Thai Tran
1
1
Faculty of Engineering and Information Technology, University of Technology, Sydney, Australia
2
Department of information technology, University of Economics, Prague, w. Churchill sq. 4, Prague 3, Czech Republic
3
Unicorn College, V Kapslovně 2767/2,130 00 Prague 3, Czech Republic
Keywords: Cloud Computing, Service-oriented Architecture, Service Life-cycle.
Abstract: As the adoption of cloud computing gathers momentum, many organizations are facing new challenges that
relate to the management of cloud computing environments that may involve hundreds of autonomous cloud
services provided by a large number of independent service providers. In this paper we argue that the large-
scale use of externally provided cloud services in enterprise applications necessitates re-assessment of the
SOA paradigm. The main contribution of this paper is the identification of the differences between service
provider and service consumer SDLC cycles and the description of the service consumer SDLC phases.
1 INTRODUCTION
Over the last decade, SOA (Service Oriented
Architecture) has become the architecture of choice
for most large organizations, providing a flexible
and responsive enterprise computing architecture
that addresses the business needs of modern
organizations. Large-scale adoption of SOA has
been facilitated by wide industry support and high
level of web services standardization resulting in
comprehensive tools and methods that support the
entire SDLC (Systems Development Life Cycle) of
service-oriented enterprise applications. A key
benefit of SOA is that it enables close alignment of
IT (Information Technology) with the business
objectives of the organization, and at the same time
facilitates high-levels of business process
automation by using discrete, loosely coupled
services. Another important benefit of SOA is that it
enables organizations to participate in the emerging
world of cloud computing and consume software
and hardware services provided by other parties.
However, the recent shift towards the use of
externally provided cloud services in enterprise
applications has altered the basic premise of SOA
and requires a re-assessment of the SOA paradigm
and in particular the SOA SDLC. The underlying
assumption of traditional IT architectures, including
SOA, is that application services are developed and
deployed on-premise by the end-user organization.
More specifically, traditional SOA assumes that the
life cycle of enterprise services is controlled by the
end-user organization, i.e. that most of the services
are designed, developed and provisioned by the
same organization that deploys the services in its
enterprise applications. Although, SOA services can
be consumed over the Internet or a private network
from remote locations (e.g. using SOAP or REST
protocols), the primary focus of traditional SOA is
support for on-premise application services. This
model is no longer valid in situations where a
significant part of enterprise systems is delivered in
the form of cloud services with a large number of
cloud service providers involved (Joshi et al., 2009).
With increasing adoption of cloud computing
(Khajeh-Hosseini et al., 2010), the primary role of
the IT consumer (end-user) organizations is shifting
from on-premise implementation of SOA
applications to integration of cloud services and
management of a hybrid environment that involves
both on-premise and cloud services (Farrell, 2011).
Public cloud services are hosted on the provider
infrastructure and controlled and managed by
external service providers that make the services
available to consumers on a pay-per-use basis
(Breiter and Behrendt, 2009). Service consumers do
not have full control over the data and applications
that are provided as cloud services, and
consequently cannot guarantee the level of security
and availability that they are typically expected to
provide (Breiter and Behrendt, 2009). In response to
such concerns, there has been significant interest in
366
Feuerlicht G. and Thai Tran H..
Adapting Service Development Life-cycle for Cloud.
DOI: 10.5220/0005470803660371
In Proceedings of the 17th International Conference on Enterprise Information Systems (ICEIS-2015), pages 366-371
ISBN: 978-989-758-098-7
Copyright
c
2015 SCITEPRESS (Science and Technology Publications, Lda.)
hybrid cloud architectures where enterprise
applications are partly hosted on-premise, and partly
in public cloud environments (Hajjat et al., 2010).
These developments impact on SOA SDLC as well
as on SOA architectural features and functions,
resulting in divergence between service consumer
and service provider SDLC cycles. As a direct
consequence of providing cloud services to large
populations of service consumers, cloud service
providers cannot take into account the needs of
individual service consumers and their service
development schedules. Although the investigation
of architectural requirements for cloud computing
has been the subject of recent research interest and
standardization efforts (Liu et al., 2011), most of the
work takes the perspective of service providers, and
relatively little research has been done so far on
issues that impact service consumers. The detail
definition of service consumer SDLC and
methodological support for cloud service consumers
are still an open research problems. In our previous
work, we discuss the challenges of managing
enterprise applications in the cloud context
(Feuerlicht and Tran, 2014a) and describe a
consumer-side solution in the form of a Service
Consumer Framework (SCF) (Feuerlicht and Tran,
2014b). In this paper, our focus is on service
consumer SDLC; we specify service consumer
SDLC phases and describe architectural components
required to support the lifecycle activities. The next
section (section 2) is a review of related work, and
section 3 is a discussion of the traditional SOA
SDLC from a service provider perspective. Section 4
describes service consumer SDLC referring to
architectural components that are required to support
this SDLC. Section 5 contains our conclusions and
proposals for future work.
2 RELATED WORK
The NIST (National Institute of Standards and
Technology) Cloud Computing Standards Roadmap,
Hogan et al. (2011) separate the role of service
provider and service consumer and define a set of
cloud computing system development life cycle
activities and functions. The activities of cloud
providers include service deployment, service
orchestration, cloud service management, security
and privacy. Activities of cloud consumers include
identifying (suitable) services, requesting services,
negotiating service contracts, and deploying
services. In early research, Papazoglou (2008)
proposes a service development lifecycle that
contains a preparation planning phase as and five
life cycle phases: analysis and design, construction
and testing, provisioning, development, and
execution and monitoring. The paper describes a
methodology and roadmap to assist service
providers and service aggregators in assembling
multiparty business processes. In a more recent
paper, Papazoglou et al. (2011) propose a change-
oriented service life-cycle to address issues that arise
with changes that cascade across multiple services.
The life-cycle starts with the identification of the
need for service change and scoping its extent, and
then progresses to a service analysis phase that uses
the model of the current state of the services (as-is
model) and the to-be service model to perform gap
analysis. Following the analysis of the impact of the
required changes, decisions are made about how to
deal with overlapping and conflicting service
functionality. During the final change life-cycle
phase new services are aligned, integrated, tested
and released into production. Using a simple travel
service scenario, Ruz et al. (2011) describe a flexible
SOA cloud life cycle using SCA (Service
Component Architecture) as a model for managing
the life cycle of service-based applications. The life
cycle consists of three phases: initial design and
deployment, runtime monitoring, and design
modification and reconfiguration. Authors also
introduce an integrated and open framework for
supporting flexible cloud service management based
on SOA principles. Gu and Lago (2007) present a
three phase stakeholder-driven service life cycle that
involves three separate roles: service provider,
service consumer and service broker. The main
objective of this approach is to decouple the
activities of service consumer and service provider
across development phases: design time, runtime,
and change time. In this stakeholder-driven life
cycle model, service provider is responsible for
service design, service development and testing, and
service consumer is responsible for service
orchestration, negotiation and monitoring. Focusing
on SDLC of cloud service, Breiter and Behrendt
(2009) describe a service lifecycle and study the
relationship of managing this life cycle and ITIL.
This approach focuses on managing IT functionality
as one or more aggregated resources exposed as a
cloud service. Another approach to managing an
integrated lifecycle of IT services in a cloud
environment is proposed by Joshi et al. (2009).
Authors propose cloud service lifecycle that consist
of five sequential phases: requirements, discovery,
negotiation, composition, and consumption. The
authors have identified performance metrics
AdaptingServiceDevelopmentLife-cycleforCloud
367
associated with each phase: data quality, cost,
security, service gap, SLA (Service Level
Agreement), QoS (Quality of Service), consumer
satisfaction, etc. Pot’vin, et al. (Pot'vin et al., 2013)
present a cloud service life cycle that aims to deliver
greater adaptability for dynamic business needs,
significant operational efficiencies, and lower
overall costs. The authors argue that to achieve the
best value from the cloud infrastructure, it is
important to have end-to-end management and
automated workflows for various activities during
all the phases of the cloud life cycle; starting with
planning and setting up the cloud, to end-user self-
service provisioning and de-provisioning of
applications, to metering and charging back for the
cloud resource usage. While the authors
acknowledge that there are multiple stakeholders
related to the service lifecycle, they apply Oracle
Enterprise Manager 12c Cloud Control (EMC12c) as
integrated business and software service lifecycle
framework that treats cloud SDLC from the provider
perspective. To support the full lifecycle from
request to retirement, Farrell (2011) introduces a
cloud lifecycle management that aims to ensure
successful use of the cloud by implementing policy-
driven provisioning processes through a self-service
portal supported by a service catalogue. The cloud
lifecycle starts with service request submitted by a
consumer using self-service portal, and then goes
through three different steps: orchestration and
provisioning, operations and governance, and
service retirement.
Most of the research work on cloud service
lifecycle reviewed in this section focuses on
provider-side SDLC and proposes methods and
procedures that aim to assist providers (or service
brokers) to successfully deliver cloud services to
consumers. This neglects to address important issues
that service consumers face in cloud computing
environments. Our focus in this paper is on service
consumer SDLC that has its own distinct phases and
characteristics that substantially diverge from
provider service SDLC. By explicitly identifying the
differences between service provider and consumer
SDLC cycles, and describing the consumer SDLC
phases we aim to develop a lifecycle methodology to
support cloud service consumers.
3 SERVICE PROVIDER SDLC
Before we discuss the specifics of service consumer
SDLC in the next section (section 4), we briefly
review service provider SDLC in this section. While
there are some differences in various approaches,
most researchers include the lifecycle phases
illustrated in Figure 1. The figure represents an
iterative and incremental lifecycle process that
includes a preparatory Planning phase and five
distinct main phases: Analysis and Design,
Implementation, Provisioning, Deployment, and
Execution and Monitoring (Papazoglou, 2008,
Papazoglou and Heuvel, 2007).
The planning phase is a preparatory phase that
analyses the business case for different combinations
of development approaches and realization
strategies. This business case analysis includes gap
analysis, and risk analysis. Gap analysis compares
planned services with available software services
that can be assembled to implement the new
business processes. Gap analysis matches high-level
descriptions of new services that make up a business
process against the available services and includes
scenario analysis that considers costs, risks, benefits,
and ROI (Return on Investment) associated with the
development of new business processes. Risk
analysis compares costs and benefits in provisioning
scenarios and verifies the feasibility of specific
options. The impact and probability of events that
may influence the performance of services is
estimated. Finally, ROI is calculated for each option
based on net benefits and costs.
Figure 1: Service Provider SDLC.
The analysis and design phase aims to identify and
conceptualize business processes as a set of
interacting services. Service analysis captures all
activities required for the identification and
contextualisation of a service, and helps to prioritize
business processes that offer potential improvements
in business value. Services are scoped aggregating
simple services and decomposing complex services
to reflect business requirements. This step is
followed by service design, during which the
ICEIS2015-17thInternationalConferenceonEnterpriseInformationSystems
368
technical details of service interfaces are specified
using design principles that include minimization of
coupling and maximization of service cohesion.
Importantly, this phase involves decisions about
service granularity and grouping of service
operations ensuring that the overlap between the
functionality of different services is minimized. The
emphasis is on maximising the potential for reuse of
resulting services in different application contexts.
The design activities are dependent on the specific
type of service and may vary according to a delivery
strategy.
Service design forms the input for the
implementation phase and includes all activities that
are related to the realisation of the services based on
the detail design developed during the previous
phase. As services are reused in different application
scenarios services need to be exhaustively tested
before they can be published in a service repository.
Testing involves ensuring that requirements as
specified in previous SDLC phase have been met
and that the deliverables are of acceptable quality
and conform to the relevant industry standards
(Kohlborn et al., 2009). Service provisioning phase
involves implementing service governance,
certification, auditing, metering, and billing, and
controlling the behaviour of services during their
use.
Service provisioning can be local or over a
network and can involve a complex mixture of
technical and business aspects that support various
client activities (Farrell, 2011).
The deployment phase involves registering
services in various service marketplaces and
repositories, determination of access rights, pricing
models, and specifying details of the corresponding
SLAs.
Service execution and monitoring phase covers
activities during service runtime when services are
operational, and their progress can be monitored.
This phase includes outage management with the
objective of maximizing availability of services.
Some methodologies (e.g. ITIL) include an
additional service improvement phase that involves
service revisions and result in the creation of new
and improved service versions.
4 SERVICE CONSUMER SDLC
As noted earlier, traditional SOA does not explicitly
differentiate between service provider and service
consumer SDLC cycles, and assumes that the
services are designed, developed and provisioned by
the same organization that deploys the services in its
on-premise enterprise applications. This assumption
is valid in most traditional SOA environments, but
does not hold in situation where enterprise
applications consume a large number of cloud
services provided by external providers with
autonomous service lifecycles. In the context of
cloud computing, service providers and service
consumers are separate entities that perform
different tasks throughout their SDLC cycles.
Service providers are responsible for the
implementation and reliable operation of services
they provide, while service consumers are primarily
responsible for the selection of suitable services,
integration of cloud services into their enterprise
applications, and ensuring continuity of operation at
runtime. We have identified the following five
phases of service consumer SDLC: Requirements
Specification, Service Identification, Service
Integration, Service Monitoring, and Service
Optimization (Figure 2). We classify these phases
into design-time activities that include requirements
specification, service identification, and service
integration, and run-time activities that involve
service monitoring, and service optimization. The
service consumer SDLC is closely interrelated with
the Service Consumer Framework (SCF) that
provides support for lifecycle phases and activities
and consist of Service Repository, Workflow
Engine, Service Adaptors, and a Notification Centre
(Feuerlicht and Tran, 2014b).
Figure 2: Service Consumer SDLC.
During the service requirements specification phase,
the service consumer describes functional and non-
functional requirements that a given service needs to
fulfil. Functional specifications of the service
describe what functions the service should provide.
While there are differences in the specification
AdaptingServiceDevelopmentLife-cycleforCloud
369
according to the type of service (e.g. application
service, infrastructure service, etc.), typically the
specification includes technical details of the service
interface (e.g. WSDL interface) and may also
include details of the technological environment
(e.g. specific hardware platforms, programming
languages, etc. in the case of infrastructure and
platform services). The non-functional attributes
include service availability, response time, and
security requirements, and may also include
requirements regarding data location, security
certification and the maximum cost of the service.
Once the service is fully described and classified, the
service consumer may create a Request for Service
(RFS) and record the information in the service
repository (Joshi et al., 2009).
Service Identification (discovery) is constrained
by the functional and non-functional requirements
documented in the previous phase (requirements
specification phase). Service identification phase
begins by searching the service consumer repository
to attempt to match the service requirements with
services that are already recorded in the repository
and certified for use. If no existing service matches
the requirements, the service consumer will search
the service repositories of cloud service providers, or
contact a preferred service provider directly to locate
a suitable cloud service. Service selection involves
identification of suitable cloud services, their testing
and approval. Service approval is an internal
certification process that certifies cloud services for
use in enterprise applications within the
organization. Given the large number of available
cloud services, the selection of suitable services can
be time consuming, in particular if this task is
performed multiple times in the context of different
projects that require similar services. Using the
consumer service repository to store information
about approved cloud services ensures that services
are shared among different projects and that the
service selection and approval process is not
unnecessarily repeated. In some instances, the
consumer may be able to negotiate details of the
SLA with the service provider, although this will
depend on the type and volume of services involved.
Following service identification phase cloud
services need to be integrated into consumer
enterprise applications. This process consists of the
registration of the application and design of
workflows using certified services already available
in the repository. Workflow design involves
searching the service repository for approved
services that match the requirements of enterprise
applications and composing workflows that control
the service execution sequence at runtime. Designers
match the desired QoS attributes values against
information stored in the repository and define the
processing rules that determine the sequence of
service execution at run-time (Feuerlicht and Tran,
2014b). This phase involves measuring QoS
attributes and comparing their values with those
specified in the corresponding SLAs. Typically, both
the service provider and service consumer perform
service monitoring independently, and both parties
are responsible for resolving service quality issues
that may arise. The SCF incorporates a Notification
Centre that records service status of cloud services
in a runtime log and is used by application
administrators to monitor service utilization, planned
maintenance activities, and to perform statistical
analysis of response times and throughput for
individual cloud services.
Service optimization phase is concerned with
continuous service improvement. This can be done
by replacing existing services with new versions as
these become available, or by identifying alternative
cloud services from a different provider with
identical functionality. For example, the payment
service PayPal could be replaced by the SecurePay
service. Using service utilization data recorded in
the SCF log and information available from cloud
service providers (or service brokers), the consumer
makes decisions about replacing existing services
based on cost and performance, and other relevant
QoS parameters. SCF supports this process of
service optimization allowing service replacement
without impacting on existing enterprise
applications. In addition to optimizing individual
services entire business processes can be optimized
by redesigning corresponding workflows.
5 CONCLUSIONS
In this paper, we have argued that the recent shift
towards large-scale use of externally provided public
cloud services in enterprise applications has altered
the basic premise of SOA, and that this necessitates
re-assessment of the SOA paradigm. In particular,
the assumption that the life cycle of enterprise
services is controlled by the end-user organization
(service consumer) is no longer valid. The
traditional SOA model that focuses on on-premise
application services is no longer sustainable in
situations where a significant part of enterprise
infrastructure and applications is delivered in the
form of cloud services with a large number of cloud
service providers involved. In our previous work, we
ICEIS2015-17thInternationalConferenceonEnterpriseInformationSystems
370
have described a proposal for consumer-side
solution in the form of a Service Consumer
Framework (SCF) that proposes architectural
extensions designed to support operation in cloud
computing environments. In this paper, we have
identified the differences between service provider
and service consumer SDLC cycles, and we have
provided high-level description of the service
consumer SDLC. Our future efforts will focus on
refining the task descriptions within the various
service consumer SDLC phases, and developing a
set of corresponding principles and guidelines that
will provide methodological support for cloud
service consumers.
REFERENCES
Breiter, G. & Behrendt, M., 2009. Life cycle and
characteristics of services in the world of cloud
computing. IBM Journal of Research and
Development, 53, 3:1-3:8.
Farrell, K., 2011. Cloud Lifecycle Management:
Managing Cloud Services from Request to Retirement.
BMC Software.
Feuerlicht, G. & Tran, H. T., 2014a. Enterprise
Application Management in Cloud Computing
Context. In The 8th International Conference on
Research and Pratical Issues of Enterprise
Information Systems (CONFENIS), Ha Noi, Vietnam.
ACM.
Feuerlicht, G. & Tran, H. T., 2014b. Service consumer
framework: Managing Service Evolution from a
Consumer Perspective. In ICEIS-2014. 16th
International Conference on Enterprise Information
Systems, Portugal. Springer.
Gu, Q. & Lago, P., 2007. A stakeholder-driven service life
cycle model for SOA. In 2nd international workshop
on Service oriented software engineering: in
conjunction with the 6th ESEC/FSE joint meeting.
ACM, 1-7.
Hajjat, M., Sun, X., Sung, Y.-W. E., Maltz, D., Rao, S.,
Sripanidkulchai, K. & Tawarmalani, M., 2010.
Cloudward bound: planning for beneficial migration
of enterprise applications to the cloud. SIGCOMM
Comput. Commun. Rev., 41, 243-254.
Hogan, M., Liu, F., Sokol, A. & Tong, J., 2011. NIST
cloud computing standards roadmap. NIST Special
Publication, 35.
Joshi, K., Finin, T. & Yesha, Y., 2009. Integrated lifecycle
of IT services in a cloud environment. In Proceedings
of The Third International Conference on the Virtual
Computing Initiative (ICVCI 2009), Research
Triangle Park, NC.
Khajeh-Hosseini, A., Greenwood, D. & Sommerville, I.,
2010. Cloud migration: A case study of migrating an
enterprise it system to iaas. In Cloud Computing
(CLOUD), 2010 IEEE 3rd International Conference
on. IEEE, 450-457.
Kohlborn, T., Korthaus, A. & Rosemann, M., 2009.
Business and Software Service Lifecycle
Management. In Enterprise Distributed Object
Computing Conference, 2009. EDOC '09. IEEE
International. 87-96.
Liu, F., Tong, J., Mao, J., Bohn, R., Messina, J., Badger,
L. & Leaf, D., 2011. NIST cloud computing reference
architecture. NIST Special Publication, 500, 292.
Papazoglou, M. 2008. Web Services Development
Lifecycle. Web services: principles and technology.
Pearson Education.
Papazoglou, M. P., Andrikopoulos, V. & Benbernou, S.,
2011. Managing Evolving Services. IEEE Software,
28, 49-55.
Papazoglou, M. P. & Heuvel, W.-J. v. d., 2007. Business
process development life cycle methodology.
Commun. ACM, 50,
79-85.
Pot'vin, K., Akela, A., Atil, G., Curtis, B., Gorbachev, A.,
Litchfield, N., Nelson, L. & Sharman, P. 2013. Cloud
Lifecycle Management. Expert Oracle Enterprise
Manager 12c. Apress.
Ruz, C., Baude, F., Sauvan, B., Mos, A. & Boulze, A.,
2011. Flexible SOA Lifecycle on the Cloud Using
SCA. In Enterprise Distributed Object Computing
Conference Workshops (EDOCW), 2011 15th IEEE
International. 275-282.
AdaptingServiceDevelopmentLife-cycleforCloud
371
