Professional Documents
Culture Documents
However, a cloud strategy can create • Industry best practices • IT strategy and governance policies
significant stress on an enterprise shifting to
• Application profile analysis • IT operations analysis
such a dynamic, user driven IT environment.
This stress can be identified and mitigated • IT infrastructure analysis • Security policies
prior to the move by conducting a cross-
• Datacenter design and analysis • Audit requirements
functional evaluation of cloud opportunities
and impact.
Cloud categories To minimize the exposure of the enterprise to • Because enterprise data can be distributed
Clouds can be categorized as public, private, the risks of cloud computing in general, and in many locations, it is difficult to achieve
or hybrid: information mismanagement in particular, the inter-operability, scalability, and
you need to implement an ongoing effort to sufficient network bandwidth to support
• A public cloud provides a flexible, educate users and track activities. adequate service levels for computing on
pay-per-use IT environment that helps the cloud.
the business to better match actual IT A few examples of the challenges facing • Implementing high-resolution monitoring
demand to expenditure, to decrease the enterprise when implementing a cloud and metering of IT infrastructure and using
capital outlay for IT infrastructure, and computing strategy are: the resulting data presents challenges
to shift IT investment and risk to a third at the technical, operational, and
party. In addition, users typically realize • Distributing data and sharing physical organizational levels.
access to greater IT functionality along resources in multiple locations creates IT
with automatic, faster control over the governance issues relating to regulatory Challenges unique to the public cloud
deployment of IT resources to support their compliance, security, and privacy. While public clouds are similar to private
work. • Many software vendors have not adapted clouds from a technological perspective,
• A private cloud operates within an their licensing models to cloud computing, they create unique challenges due to their
enterprise datacenter or in an externally making it difficult for the enterprise to nature as remote, shared entities. These
hosted datacenter for the sole benefit of a ensure licensing compliance in a cloud include dependency on external suppliers
single enterprise. A private cloud is easier environment. for business-critical infrastructure — with an
to align with security, compliance, and associated loss of control — and the use of a
regulatory requirements, and provides
more enterprise control over deployment
and use.
• A hybrid cloud delivers IT through a mix of
both public and private clouds. Cloud End Users
shared environment, which includes potential The benefits of cloud computing are less Cloud deployment models
availability, latency, security, privacy, and pronounced where applications are constant In some cases, for technical, operational,
regulatory issues. in their resource demands, are not easy to or business reasons, a given application is
partition for dynamic provisioning, or do not concurrently deployed on traditional non-
Legal issues in the public cloud require a high level of scalability. In these cloud infrastructure and on a cloud, or on a
Data storage in the public cloud creates environments, the risk and cost associated hybrid private/public cloud. Several general
additional legal and regulatory challenges. with transitioning to the cloud can be more models for these types of deployment are
For example, information stored on a cloud difficult to justify. recognized:
should be considered encapsulated and
therefore immune from government access, The type of cloud that is most suitable for • Test and development — where the needs
unless appropriate court orders are obtained. a given enterprise — public, private, or of development, testing, and early stage
Unfortunately, there is concern that the public hybrid — and the service layers required — deployment benefit from low overhead,
cloud owner could grant government access infrastructure, platform, or software — often rapid provisioning, and the wide variety of
to the information without the knowledge or depend on the types of applications the software and hardware resources typically
consent of its owners. enterprise commonly uses. available on the cloud
• Functional off-load — where certain, well
In general, the legal issues of computing in Applications that should move to the cloud defined compute or storage functions are
the public cloud are still unclear and only Applications should move to the cloud based moved to the cloud, and other functions
time will tell if adequate solutions can be on the cost/benefit considerations of the are retained in non-cloud environments
found to alleviate the issues without creating move. Typically, it is easy to justify moving • Augmentation — where the public cloud
barriers to deployment. applications that can benefit from rapid is used to service peaks in demand in a
deployment, need a high level of elasticity, private datacenter
Industry-specific considerations and can scale easily. Applications that are not • Web services — where a well defined
The justifications, issues, and benefits of standards based, or that are dependent on service that is provided to many
transitioning to the cloud differ between software that is not supported on the cloud, applications can benefit from the elastic
industries and applications types. can be difficult to move. When transitioning nature of the cloud to service different
For example, in financial services, over- to the cloud, you can choose to replace levels of demand
provisioning is the most common strategy these difficult-to-transition applications or
used to deal with demand spikes, despite the to maintain them in a traditional, non-cloud Route to the cloud
high cost and inefficiency of this approach. environment. In any case, storage, backup, or Whether the target deployment is to a public
If the enterprise uses cloud computing, it networking can still be provided by an IaaS cloud, private cloud (in-house or hosted), or
can service demand peaks by allocating cloud to applications that cannot transition to hybrid, moving an enterprise’s IT to the cloud
datacenter capacity to handle them and, the cloud in full. is a complex task that requires fundamental
once the demand peak is over, the datacenter changes to many IT related processes. To help
capacity can be applied to other, lower If considering a move to a multitenant cloud, ensure a successful transition, the route to
priority work loads, thus driving up average whether public or private, applications that the cloud typically consists of the following
utilization. require security but cannot be secured, sequential stages, as illustrated in Figure 3.
cannot be moved. For an expanded discussion
Where large compute and storage farms of security on the cloud, see “Security on the When considering such a transition, it is
are needed to support highly scalable, cloud” on page 8. important to remember that cloud computing
dynamic workloads such as those typical is an evolution of current datacenter
of HPC applications or publicly facing Web practices, architectures, and technologies.
applications, the business justifications of Over the last few years, your organization
the cloud are immediate. In fact, many of the might have taken steps to rationalize your IT
technologies that underlie cloud computing and make it more efficient, for example, by
were developed to service these types of deploying your applications on virtualized
applications. servers. If this is the case, you have already
5 Kickstart Your Transition to the Cloud
taken the first steps in your transition, which The initial analysis identifies the extent of track, manage, and plan datacenter capacity.
makes it much simpler to fully transition to these challenges in the enterprise so that the The end result is that the datacenter as a
the cloud. following stages can focus on resolving the whole can reach operational and architectural
challenges in the course of the transition. maturity.
Analyzing the present state
Existing enterprise IT is often divided into Virtualizing, consolidating, and retiring Transitioning to the cloud
several organizational and technological equipment Finally, the last step in evolving the
silos. The fact that business units are usually On the cloud, virtual machines are the datacenter to the cloud can begin by
the owners of applications can result in an standard deployment object. Virtualization implementing the various components of the
environment where there is minimal sharing enables flexibility by abstracting the cloud infrastructure in phases, including:
of resources. This creates several issues and hardware. Software can be deployed without
challenges: depending on any specific hardware resource. • Physical resource virtualization
Once the applications are virtualized, • Self provisioning
• Widely diverging standards or proprietary the datacenter can be rationalized and • Programmatic control
interfaces — each business unit is free to consolidated at the infrastructure, platform, • Dynamic resource allocation
select its own technological components and application levels, and superfluous • Pay-per-use
• Minimal interaction between different equipment can be retired.
applications and minimal automatic External interfaces, often required to enable
information sharing — divergent Optimizing operations information and resource sharing between
development creates difficulties in Decoupling applications from hardware helps loosely coupled clouds, are implemented as
application integration enable facility and IT operations integration, part of the transition.
• Poor utilization — business units tends to infrastructure sharing, and improved
maintain their own, over-provisioned, spare automation. These in turn all contribute to
capacity optimizing resource usage and achieving
• Diverse operations and processes — the SLA targets. The metering capabilities of
needs of the business units differ and there cloud computing create extensive, high
is no entity within the enterprise that resolution usage data across all aspects of the
enforces common processes datacenter operations. In addition to enabling
pay-per-use, the data collected can be used to
Consolidation,
Optimization in
Silos Virtualization, and Cloud
Operations
Retirement
Figure 3. Organizations should evolve from a state of multiple, disjoint silos with their associated inefficiencies, through increasingly advanced
operational models to the cloud.
6 Kickstart Your Transition to the Cloud
Strategic planning • Enhanced business continuity and disaster requirements are examined and validated
While cloud computing focuses primarily on recovery capabilities and aspects of the current deployment that
solving the issues associated with enterprise • Efficiency of IT datacenter deployment might inhibit progress are identified, together
IT, it can deliver additional benefits by helping • Testing and handling of demand peaks with aspects that are likely to support it. The
to create an agile, adaptive enterprise. The • Integration with other organizations when set of validated requirements are assessed to
Sun Cloud Strategic Planning Service (Figure required by mergers and acquisitions identify critical success factors and perform a
4) helps you implement a strategic analysis gap analysis.
and leverage the opportunities created by the Sun’s cloud implementation methodology
move to the cloud. The strategic planning and To transition successfully to the cloud, you Synthesis
analysis phase is the first stage of Sun’s cloud need to implement a structured, multistaged The synthesis phase maps the requirements
services engagement. Sun cloud professionals approach to plan, control, and reduce the and target deployment models to appropriate
apply a set of industry best practices and risk of the transition. Sun’s proven SunScope architectural options. This information is then
evaluation techniques to analyze the current process (Figure 5) helps you transition to the used to design alternative solution options.
enterprise environment and identify cloud cloud through the following stages: The alternative solution options are then
computing opportunities. Each of the evaluated to help ensure they can meet the
elements listed in the following sections are • Discovery requirements and remedy the gaps, and the
evaluated in the context of the needs and • Synthesis target solutions are defined. Finally, solution
challenges of the enterprise, and the potential • Design deployment programs plans are created and
benefit to these elements that can be derived • Execute, realize prepared for the design phase.
from cloud computing. • Sustain
n/
Cul
Bus
making
tio
Tec
pe
h
lo g
• Employee productivity, collaboration, and y IT O
innovation
Design to the actual transition, since the applications technologies, and operational models
The detailed design phase creates the designs are normally backward compatible with that can help resolve the inefficiencies of
for the target architectures as derived from traditional platforms. The deployment to enterprise IT. Cloud computing features
the synthesized program plans. Some of the cloud is implemented after application — elasticity, resource virtualization, self
the assumptions and designs might require changes are fully verified. provisioning, metering, multitenancy, and
confirmation — by implementing proofs-of- more — provide you with the tools you need
concept, benchmarks, or pilot deployments — Sustain to deploy IT services with unparalleled levels
to finalize the design of the different elements After successfully deploying the appropriate of efficiency. However, successfully adopting
and help ensure they match the requirements. elements of enterprise IT to the cloud, the cloud computing requires that you transform
The products of the design phase are concrete transition is complete and the organization your IT and implement fundamental
operational and engineering implementation moves its cloud operations into sustaining organizational, technological, and operational
plans. mode. In this stage, the enterprise changes.
implements the maintenance plans created
Execute, realize in the design phase, including the operational Sun has been supporting the adoption of
The execution phase begins by validating support structure. Now you can monitor the cloud computing in the enterprise, building
the plans and reviewing the expectations performance of your IT operations and the key cloud architectures, technologies for cloud
of the different stakeholders regarding the IT performance indicators, and validate the computing, and providing public cloud
implementation process. The transition attainment of your operational objectives. services for several years. Sun Professional
is executed with Sun’s standard project Services personnel have gained a wealth of
management methodologies, which maintain Kickstart your transition to the cloud with Sun experience in helping enterprises address
ongoing contact with stakeholders and Services the challenges of cloud computing. Sun’s
deliver planned milestones. The physical Enterprise IT is facing increasing pressure service offerings include strategic and tactical
systems are deployed in phases, with different to become more efficient. This pressure is analysis, evaluating the current environment
applications transitioning to the cloud driven by the difficult economic environment, and cloud opportunities, planning the
gradually to help ensure a smooth transition increased competition, and regulatory transition to the cloud, and managed services
with minimal disruption. It is important to pressure to control carbon emissions. The IT to support you in adopting and sustaining
remember that the applications themselves industry is responding to this challenge by cloud computing.
can be adapted to the cloud and tested prior developing new and innovative architectures,
• Define and validate • Analyze requirements • Detailed design of • Validate implementation • Monitor and manage
requirements and architecture options candidate architectures plans and set solutions
• Identify client enablers and • Design solution options • Engineer solutions expectations • Monitor key performance
inhibitors • Evaluate solutions • Implementation indicators
• Define target state vision and • Create plans for • Monitor implementation • Publish IT health reports
success factors deployment and gap • Milestone checkpoints • Institutionalize IT
• Perform gap analysis remediation with stakeholders management policies
Key Results
Figure 5. Sun’s proven SunScope methodology helps you through the different stages of moving to the cloud.
8 Kickstart Your Transition to the Cloud
Sun Microsystems, Inc. 4150 Network Circle, Santa Clara, CA 95054 USA Phone 1-650-960-1300 or 1-800-555-9SUN Web sun.com
© 2009 Sun Microsystems, Inc. All rights reserved. Sun, Sun Microsystems, and the Sun logo are trademarks or registered trademarks of Sun Microsystems, Inc., or its subsidiaries in the United States and other countries. Information
subject to change without notice. SunWIN #568989 Lit. #SVSB15023-0 09/09