Dispelling The Vapor Around Cloud Computing –Network Management

IBM Smart Business January 2010
Thought Leadership White Paper

Dispelling the vapor
around cloud computing
Drivers, barriers and considerations for public and
private cloud adoption

2 Dispelling the vapor around cloud computing

Before developing a cloud strategy, it’s important to under-
Contents stand what cloud computing really is. Cloud computing is
2 Introduction both a user experience and a business model. It is an emerging
style of computing in which applications, data and IT
3 Survey scope and definitions resources are provided to users as services delivered over the
network. It enables self-service, economies of scale and flexible
5 Cloud computing adoption and consideration factors
sourcing options. In addition, cloud computing is an infra-
6 Workload preferences for private and public cloud structure management methodology—a way of managing
large numbers of highly virtualized resources, which can reside
10 Service management process criticality in multiple locations, so that they resemble a single large
12 Key success factors resource that can be used to deliver services. Additionally,
there are multiple cloud delivery models available—some are
13 The steps to cloud adoption internal (private), some external (public), and some are a com-
bination of both (hybrid).
Introduction
In recent years, cloud computing environments have been To better understand the current rate of adoption, as well as
gaining popularity. In the IBM Global CIO Study 2009, when drivers, barriers, and considerations that are influencing the
asked about their most important visionary initiatives, cloud adoption of cloud computing, IBM conducted a survey in June
computing was cited by more than one-third of CIOs as being and July of 2009 of 1,090 IT and line-of-business decision
among the most important. makers around the world. The purpose of this white paper is
to share the survey findings, and to provide the IBM point of
The many inherent benefits of a cloud environment make it view regarding key steps and considerations for cloud adop-
attractive to organizations looking to speed service delivery tion. The findings validate that many organizations are at least
and increase IT efficiencies while supporting information considering cloud computing—most favor a private cloud
management, service management and service-oriented archi- delivery model at this time—yet significantly, less than a quar-
tecture (SOA) initiatives. While cloud computing offers dis- ter of organizations surveyed have mature service management
tinct advantages, organizations that are considering adoption and infrastructure capabilities that are essential for cloud.
must carefully strategize about delivery models, workloads and
infrastructure readiness.

IBM Smart Business 3

Survey scope and definitions The survey addressed three aspects of cloud computing:
The popularity of cloud computing and its potential to trans-
form service delivery has led a growing number of organiza- ● Sourcing choices—adoption or consideration of cloud delivery
tions to implement cloud computing for one or more models, plus drivers behind and barriers to adoption
workloads, while many others are actively considering it. ● Workload considerations—public and private cloud preferences
Those surveyed included IT and line-of-business decision by workload type
makers across the U.S., Canada, France, Italy, Germany, the ● Service delivery within the cloud—service management consid-
UK, Japan, China and India. Most were from companies of erations and infrastructure readiness
between 1,000 and 9,999 employees (see Figure 1).
Respondents represented a variety of industries, including Survey definitions
public sector, communications, distribution, financial services, Clouds were defined according to three types: public, private
industrial and others. and hybrid.

Public clouds—IT activities/functions are provided “as a
Job role Country service,” over the Internet, which allows access to technology-
C-level
Business India enabled services without knowledge of, expertise with, or con-
Business
Executives
2%
Leaders 15%
USA trol over the technology infrastructure that supports them.
30%
22% China 10%
Also called “external cloud.”

11%
62%
14% Japan
5% Private clouds—activities and functions are provided “as a
CIOs 9% 10% Canada
10% service,” over a company’s intranet. It is built by an organiza-
UK
IT
France tion for its own users, and everything is delivered within the
Germany
Directors organization’s firewall (instead of the Internet). The private
Sector cloud owner does not share resources with any other compa-
Communication
Other nies, so multitenancy is not an issue. Also called an “internal
Services
4% Distribution cloud.”
17%
22%

Public
18% Hybrid clouds—the external and internal service delivery
12%
Financial methods are integrated. Rules and policies are established by
27% Services
the organization based on factors such as security needs, criti-
Industrial
cality and underlying architecture, so that activities and tasks
are allocated to external or internal clouds as appropriate.
Source: IBM Market Insights, Cloud Computing Research, July 2009. n=1,090

Figure 1. Survey demographics. Respondents to IBM’s cloud comput-
ing adoption survey included IT and business decision makers from eight
countries, representing a cross section of industries and company sizes.


While 73 percent of respondents said they were already famil-
Workload type Workload
iar with the internal and external (private and public cloud)
delivery models described in the survey, there was little consis- Analytics ● Data mining, text mining, or
tency in the terms that respondents associated with these other analytics
● Data warehouses or data marts
delivery methods. For example, 30 percent of respondents ● Transactional databases
selected “software as a service” as the term that best described
the delivery models, while only 24 percent selected “cloud Business services ● CRM or sales force automation
● E-mail
computing.” Other choices included “hosting” (19 percent), ● ERP applications
virtualization/consolidation (16 percent) and utility computing ● Industry-speciﬁc applications
(4 percent).
Collaboration ● Audio/video/Web conferencing
● Uniﬁed communications
As part of the study, respondents were asked to rate workloads ● VoIP infrastructure
they would most consider for deployment in a public or pri-
Desktop and devices ● Desktop
vate cloud environment.
Development and test ● Development environment
● Test environment
Workloads—the kind of work that an organization needs to
accomplish. Each workload has characteristics that make it run Infrastructure ● Application servers
● Application streaming
most efficiently on certain types of hardware and software.
● Business continuity/disaster
Some demand fast transactions, like ATMs, while others, like recovery
predictive analytics, require intense calculations. The full list ● Data archiving
of 25 workloads included in the survey is shown in Table 1. ● Data backup
● Data center network capacity
● Security
● Servers
● Storage
● Training infrastructure
● WAN capacity

Source: IBM Market Insights, Cloud Computing Strategy Research,
July 2009.

Table 1: Workloads. The study asked respondents to rate 25 different
workloads they had already deployed or would consider deploying in a
public or private cloud.


Cloud computing adoption and a 20 to 29 percent cost savings to make a compelling case for a
consideration factors migration to cloud. In IBM’s experience, actual savings in
Organizations around the globe are drawn to cloud computing many cases exceed that 20 to 29 percent threshold. However,
for its ability to speed service delivery and increase service and the level of savings achievable depends upon many factors,
infrastructure availability while creating an elasticity that including the types of workloads being processed, the delivery
allows services to be expanded—or contracted—as demand method selected and the efficiency of the infrastructure to be
changes. And because organizations can leverage economies of replaced by cloud.
scale through a cloud environment, they can also reap a higher
ROI via greater staff efficiency and optimization of IT Factors that deliver faster time to value were identified as key
resources. Cloud computing supports efforts to establish a drivers by 72 percent of respondents, including relieving pres-
service-oriented architecture and to enhance information sure on internal resources, simplifying system updating or
management and service management. With all these benefits upgrading, and being able to scale IT resources to meet needs.
in place, cloud computing can also help improve the percep- And half of respondents said they were motivated by the abil-
tion of IT as being both flexible and responsive. ity to improve system availability and reliability.

The study found that decision makers express openness to
both public and private cloud, although the consideration and Improve system reliability and availability 50%

adoption rates for the private cloud delivery model were Pay only for what we use 50%

Hardware savings 47%
higher. Of those surveyed, 64 percent rated private cloud Software licenses savings 46%
delivery “very appealing or appealing,” compared to Lower labor costs 44%

38 percent for hybrid cloud and 30 percent for public cloud. Lower outside maintenance costs 42%

Reduce IT support needs 40%
What’s more, at least 25 percent of decision makers report
Able to take advantage of latest functionality 40%
that they have already implemented an internal cloud. Relieve pressure on internal resources 39%

Resolve problems related to updating/upgrading 39%

Cost savings and time to value are the leading drivers Rapid deployment 39%

Able to scale IT resources to meet needs 39%
The survey reveals the primary drivers for cloud adoption in
Able to focus on core competencies 38%
general, and public cloud adoption in particular are cost sav- Take advantage of improved economies of scale 37%
ings and faster time to value. In fact, as shown in Figure 2, Reduce infrastructure management needs 37%

77 percent of respondents chose cost savings as a key driver Lower energy costs 29%

Increase budget flexibility by transforming capex into opex 29%
for public cloud adoption, citing anticipated savings in the
Reduce space required to house/maintain infra. 26%
areas of software licenses, hardware, labor, IT support and Create new revenue streams 23%

maintenance. One-third of respondents said they require
0% 20% 40% 60%
Source: IBM Market Insights, Cloud Computing Research, July 2009.

Reduce costs Speed time to value Improve reliability Other

Figure 2. Public cloud drivers. Cost savings, time to value and
improved availability are the leading attractions for using a public cloud
service.


Data security concerns are a barrier to public cloud infrastructure and provisioning, a type of IT infrastructure
adoption workload in which IT resources are allocated to test applica-
Sixty-nine percent of respondents said that concerns about tions or systems management initiatives, such as server virtual-
data security and privacy were the primary barrier to public ization, in a protected environment so that production
cloud adoption, and more than half cited concerns about serv- activities are unaffected.
ice quality and performance (54 percent), doubts about true
cost savings (53 percent) and insufficient responsiveness over In the study, we found that workload types favored for public
the network (52 percent) as hindrances to public cloud adop- and private cloud deployment varied widely. Conferencing and
tion. Workloads rated as the least favored for public cloud, by CRM/sales force automation were among the highest ranked
far, were those related to security and databases. Despite those for usage consideration for public cloud, while they were
concerns, public cloud adoption is expected to gain momen- among the lowest for private cloud. Within public cloud rank-
tum in coming years. Though private cloud is most favored ings, there was more than a 20 percent spread between the
now, adoption of public clouds is expected to grow by highest and lowest ranked workloads. For private cloud, con-
26 percent CAGR between now and 2013.1 sideration rates were much more consistent, varying by less
than 10 percent for the top 15 workloads.
Workload preferences for private and
public cloud For this analysis, “consideration” is deﬁned by the percentage
Since workloads vary according to platform requirements, of respondents reporting that, for a particular workload, they:
complexity, business criticality and data sensitivity, some are
ideal for a public cloud, while others are more suitable for ● Had already implemented an internal or external delivery
deployment in a private cloud. For this reason, a workload- method
based approach to cloud computing is essential in determining ● Were planning an internal or external delivery method
which delivery model is best for that workload—public, within the next 12 months
private or hybrid—and in understanding which workloads can ● Would consider implementing an internal or external
realize the greatest beneﬁts in a cloud environment. delivery method in the next 12 months
● Would consider implementing an internal or external
In addition, some workloads, due to their characteristics (low delivery method in more than 12 months.
risk and high potential for ROI) are better candidates for a
cloud pilot project. An example of this is test environment


Workload preferences for private cloud
Database and application-oriented workloads emerged as most
Data mining, text mining,
appropriate for potential private cloud adopters. However, pri- or other analytics 34% 12% 16% 21% 12% 5%

vate cloud consideration was high across all workloads. As
Application streaming 31% 11% 21% 18% 16% 3%
shown in Figure 3, the top five considered workloads included
Service/help desk
data mining, text mining, or other analytics, by 83 percent; 46% 7% 13% 14% 15% 3%

application streaming, by 81 percent; service/help desk by Industry-specific 41% 8% 16% 15% 16% 4%
applications
80 percent; industry-specific applications by 80 percent; and
37% 8% 17% 18% 15% 5%
test and development environments tied at 80 percent. Test environment

Development 36% 7% 18% 19% 13% 7%
environment
Moreover, actual implementation rates for private cloud were
Application servers 44% 6% 14% 15% 15% 6%
high across all workloads, although not as consistent as con-
Training/demo
sideration rates. Forty-six percent have implemented service/ environment
35% 9% 14% 21% 15% 6%

help desk. Other workloads that nearly half of respondents Data warehouses 37% 10% 16% 15% 18% 4%
or data marts
said they had already implemented in a private cloud include
Long-term data 34% 11% 16% 17% 18%
desktop (45 percent), e-mail and application server workloads archiving/preservation
4%

(both 44 percent). Unified communications 30% 13% 20% 15% 15% 7%

Storage 42% 8% 13% 14% 17% 6%

Transactional 40% 7% 14% 16% 17% 6%
databases

Servers 43% 8% 13% 13% 17% 6%

ERP applications 40% 9% 11% 16% 19% 5%

0% 20% 40% 60% 80% 100%

Have implemented an internal delivery method
Currently plan implementation of an internal delivery method in the next 12 months
Would consider implementing an internal delivery method in the next 12 months
Would consider implementing an internal delivery method in more than 12 months
Would not consider implementing an internal delivery method
Don’t know

Figure 3. Preferred private cloud workloads. Analytics, application
streaming, service/help desk, industry-specific applications and test and
development environments were most favored for a private cloud.


Workload preferences for public cloud
Overall, infrastructure workloads emerged as most favored by Audio/video/
Web conferencing 23% 9% 19% 20% 25% 4%
respondents for public clouds, but the most attractive usage
CRM or Sales
was for CRM, an application workload. As shown in Figure 4, Force Automation 18% 11% 12% 20% 34% 5%

the top ﬁve favored workloads for public cloud deployment, Business continuity/
disaster recovery 11% 7% 18% 24% 34% 6%
based on those who said they were currently using, were plan-
Long-term data
ning to or would consider implementing them in a year or archiving/preservation 10% 7% 15% 27% 35% 6%

more, included: audio/video/Web conferencing, by 71 percent;
Application streaming 11% 7% 19% 21% 37% 5%
CRM or sales force automation, by 61 percent; business conti-
Training/demo
nuity/disaster recovery, by 60 percent; data archiving, by environment 9% 7% 17% 25% 35% 7%

59 percent; and application streaming, by 58 percent. CRM Unified
7% 10% 18% 22% 35% 8%
communications
and conferencing are already popular software as a service
WAN capacity 17% 6% 14% 18% 38% 7%
(SaaS) workloads, which makes them logical selections for
public cloud implementation. E-mail 15% 6% 16% 16% 43% 4%

Development 9% 5% 12% 26% 41% 7%
environment

Short-term data
backup and recovery 9% 5% 16% 20% 44% 6%

Industry-specific
applications 13% 6% 13% 17% 47% 4%

Service/help desk 11% 4% 15% 19% 45% 6%

VoIP infrastructure 10% 6% 13% 20% 44% 7%

Data center
network capacity 10% 4% 16% 19% 43% 8%

0% 20% 40% 60% 80% 100%

Have implemented an external delivery method
Currently plan implementation of an external delivery method in the next 12 months
Would consider implementing an external delivery method in the next 12 months
Would consider implementing an external delivery method in more than 12 months
Would not consider implementing an external delivery method
Don’t know

Figure 4. Preferred public cloud workloads. Conferencing and
CRM or sales force automation topped the list of preferred public cloud
workloads.


On the other hand, respondents have clearly indicated their IBM derived a public cloud adoption index and a private cloud
position that not all workloads are good candidates for migra- adoption index for each workload by analyzing:
tion to a public cloud environment. Some examples include:
● The implementation rankings for public versus private
● Highly sensitive data workloads (e.g. employee and health clouds
care records) ● The barriers against and push factors for private and public
● Multiple, codependent services (e.g. high throughput online cloud adoption
transaction processing) ● The relative budget expenditure.
● Workloads requiring a high level of auditability and
accountability (e.g. those subject to Sarbanes-Oxley) As discussed above, the self-reported consideration rate for
● Third-party software that does not have a virtualization or private cloud workloads was consistent, with between 40 and
cloud-aware licensing strategy 50 percent reporting current or planned private cloud imple-
● Workloads requiring detailed chargeback or utilization mentation across all workloads. However, a different picture
measurement (e.g. capacity planning and department-level emerged with the adoption index analysis, with a wider spread
billing) between the workloads with the highest push/lowest barrier
● Workloads requiring customization (e.g. customized ERP factors and those with the lowest push/highest barrier profile.
applications)
Workload recommendations
Push factors versus barriers Based on the adoption index analysis and IBM’s own cloud
As part of the survey, respondents were asked to identify the implementation experience, both internally and with clients,
factors that would influence their deployment choices for each IBM has identified the workloads that offer the most favorable
workload—either for or against public or private cloud adop- entry points for public and for private cloud delivery models.
tion. “Fluctuating demand” and “unacceptably high costs for For organizations interested in piloting a public cloud service,
traditional IT delivery” emerged as “push” factors for both the infrastructure workloads listed in Table 2 will most likely
public and private cloud. The factors that skewed preference be the projects that will pose the lowest risk and offer highest
in favor of public cloud (and against private) were whether or potential return. The same holds true for the workloads listed
not the workloads are highly standardized, modular and inde- as top candidates for private cloud implementation.
pendent. Conversely, concerns about data privacy, compliance
with regulatory requirements and being able to maintain a Note that “test environment infrastructure” appears in both
high level of control over the environment were associated lists. This is because it is a relatively low-risk workload in
with a higher propensity to choose private over public cloud terms of the business and the overall IT operation. In addi-
models. tion, this workload has the potential for large ROI through


gains in speed and cost reduction. Therefore, IBM recom- Some workloads are ideal for migration to a cloud environ-
mends that test environment infrastructure workloads should ment, while others are not, so it is essential for organizations
be considered a strong candidate for a pilot cloud project in to understand workload infrastructure and business require-
either delivery model. ments to adopt the optimum cloud environment and to fully
realize its benefits. IBM Research has developed a workload
analysis tool that can help characterize and prioritize candidate
workloads for delivery from the cloud.
Public cloud entry points: infrastructure workloads are
most appropriate
Service management process criticality
● Audio/video/Web conferencing The study also explored certain readiness factors for cloud
● Service help desk implementation. Respondents were asked to rank 18 service
● Infrastructure for training and demonstration management processes and functions based on how critical
● WAN capacity each is to the organization. The ranking was on a scale
● VoIP Infrastructure
from 1 to 5, where “1” was “not critical at all” and “5” was
● Desktop
“extremely critical.” Figure 5 shows the percentage of respon-
● Test environment infrastructure
dents who rated each function either a 4 or 5: “very critical”
● Storage
● Data center network capacity or “extremely critical.”
● Server
As shown in Figure 5, the majority of respondents said they
Private cloud entry points: database and application work- were focused on the traditional disciplines of systems manage-
loads are most appropriate ment: security, networks, servers and applications. Ranked as
less critical were the next layer of service management: avail-
● Data mining, text mining, or other analytics ability, problem, capacity and change management.
● Security
● Data warehouses or data marts
The significant finding is the gap that the survey revealed
● Business continuity and disaster recovery
between the rate of organizations that have deployed, or plan
● Test environment infrastructure
● Long-term data archiving/preservation
to deploy, a private cloud and the importance given to service
● Transactional databases management capabilities that are essential to the delivery of
● Industry-specific applications
● ERP applications

Table 2. Workload recommendations. Based on the analysis of study
data and experience with actual cloud implementations, IBM has identified
workloads that will most likely pose the lowest risk and offer highest
potential return for cloud projects.


services via cloud. In a cloud environment, service manage- Cloud computing is the ultimate step in building a dynamic
ment is key to automating self-service, automating provision- infrastructure, as discussed in the next section. In fact, the
ing and managing access and security across the environment. transformative steps between building a dynamic infrastruc-
Yet, as seen in Figure 5, these functions were rated as critical ture and progressing to a cloud computing environment are
by fewer than 25 percent of respondents. defined as the addition of provisioning, usage accounting and
self-service catalog capabilities. For these reasons, IBM rec-
ommends prioritizing the following service management
Security management 70% processes, each of which is essential to successful service deliv-
Network management 68% ery in a private cloud. For organizations considering public
Server performance and management 65% cloud services, it will be important to verify that the service
57%
Application/database performance management
provider is able to provide these capabilities.
Storage performance and management 49%

Availability management 47%
Service catalog—Self-service is a key defining capability for
Incident and problem management 43%

Capacity management 36%
cloud. The service catalog is the interface that allows
Change and configuration management 36% users/customers to select, size and order the services they
Device management 34% want.
Event management and monitoring 32%

Governance 32%
Governance—Governance is the discipline that helps
Asset management 30%
minimize risk, maximize value and align IT and business
Release management 22%
objectives.
Billing/metering 22%

Software image management 21%

Provisioning logic 12% Provisioning logic—The provisioning function works
Service catalog 11% with the service catalog by translating the user’s request for a
service—for example, a test environment with “x, y and z”
0% 20% 40% 60% 80% characteristics—into a specific infrastructure configuration.
Percent Rating As Critical (4 or 5)
Respondents could rate multiple processes as critical
Most importantly, this provisioning must take place automati-
Source: IBM Market Insights, Cloud Computing Research, July 2009. cally, without any hands-on intervention by the IT staff.
Traditional systems management disciplines
Service management for a dynamic infrastructure Usage/accounting—Also known as “metering and billing,” this
Service management for cloud
process tracks actual usage according to defined metrics and
translates them into either an internal charge-back transaction
Figure 5. Service management process criticality. Responses or a bill for the public cloud customer.
showed a lack of importance given to service management capabilities
that are essential to the delivery of services via cloud.


Overall, the survey revealed that although respondents are In addition to the gap in service management readiness, the
more interested in private cloud, there are significant gaps in survey also revealed that few respondents are far along in the
organizational and infrastructure readiness, with fewer than virtualization maturity continuum. In fact, as shown in
25 percent addressing service management issues that are Figure 6, only 18 percent indicated that virtualization is con-
essential to deliver services within a private cloud. sidered a strategic objective or a transformative technology,
while one-third reported that they are not currently using vir-
Key success factors tualization technologies at all. If not already in place, organi-
Just as service management considerations are critical to real- zations can look to IBM to help implement the capabilities
izing the benefits of a cloud computing environment, so is needed for a cloud-ready infrastructure.
establishing a dynamic infrastructure, with capabilities for
resource virtualization, service standardization and automated
provisioning of IT resources (which delivers the greatest Level 0: Do not use virtualization
33%
potential for cost savings). Implementation of private cloud is technologies at all

a best-practice evolution of a dynamic infrastructure, which
Level 1: Virtualization is only used in
has the characteristics necessary to improve the quality and pilot, proof-of-concept, or limited 14%
development deployments
speed—and reduce costs of—delivering IT services.
Level 2: Using virtualization to
replace physical development, test, 17%
Specifically, an infrastructure that is “cloud-ready” will have and/or noncritical production servers

the following characteristics and capabilities: Level 3: Our default is to use virtual
machines, unless counter-case can 12%
be made
● Open-standards-based and service-oriented Level 4: Virtualization is a strategic
objective for company 14%
● Advanced virtualization and automated management
● Common components and processes
Level 5: Virtualization is
4%
● Advanced security and resiliency. transformative to our business

Don’t know 6%

0% 20% 40%


Figure 6. Use of virtualization technology. Only 18 percent or
respondents indicated that virtualization is considered a strategic objective
or a transformative technology for their organizations.


The steps toward cloud computing cloud engagements with customers around the world and
Having a cloud adoption strategy and following the correct within IBM. Today, IBM is helping organizations around the
pathway to adoption are essential to success. This can be the world get started on the journey toward cloud computing by
most challenging phase of the adoption process, as was born following six strategic steps to cloud adoption.
out in this research, and it is the stage at which most respon-
dents said they would be prone to request outside assistance. 1. Create the IT strategy and roadmap
With a strategy that prioritizes workload types for considera- Begin by determining the goals for the organization regarding
tion and maps them to the optimal delivery model, cloud cloud service delivery, and the platform requirements and
computing presents a tremendous opportunity for IT organi- complexity associated with each. Goals can include:
zations to increase business value and ROI.
● Centralization to reduce data centers and costs
In a world that is becoming more instrumented, intercon- ● Consolidation of servers and storage to reduce the carbon
nected and intelligent, cloud computing represents a smarter footprint
way to deliver services, use resources, control costs and main- ● Virtualization to increase hardware utilization and standard-
tain a competitive advantage. With an IT strategy that ize virtual images
embraces cloud computing, CIOs can better satisfy their cus- ● Automation of service management processes application
tomers through improved visibility of IT resource use, better migration
responsiveness, simplification and cost-effective service deliv- ● Optimization of resources and dynamic provisioning of
ery. Some CIOs worry that cloud computing will be disruptive services.
to IT operations, but properly implemented, cloud computing
2. Assess and select workloads for cloud
can help lower costs, increase IT responsiveness and optimize
Identifying and prioritizing workloads that are candidates for
service delivery. Cloud computing can also increase the range
cloud delivery is the next step. Workloads that are amenable
of an organization’s services, applications and capabilities
to standardization, are self-contained applications or have a
available to customers. To make the benefits of cloud comput-
service-oriented architecture are more likely to be easily
ing a reality, CIOs must embrace the necessary changes.
ported to a cloud environment. Conversely, workloads that
will be more difficult, risky and costly to migrate to cloud
Successful implementation requires an overall strategic
computing are typically highly complex, require high amounts
vision combined with a pragmatic, evolutionary approach to
of data transfer or involve a high degree of customization or
deployment—one that IBM has developed through its many
are legacy systems that would require significant redesign to
conform to cloud architecture.


3. Determine the cloud delivery model 5. Establish the architecture
Once you have identified candidate workloads, you can then The final step prior to implementation is to establish the
establish which are more suited for delivery via a public cloud, architecture that will support your cloud initiatives, both pub-
a private cloud, or a hybrid of both. For public clouds, infra- lic and private. A cloud architecture should address three
structure and software-as-a-service workloads pose the least “domains:”
risk and offer the highest potential ROI. These include audio,
video and Web conferencing, help desk and training infra- ● What are the services you will deliver or acquire?
structure, desktop environments and storage. For private These include infrastructure as a service (“raw” computing
clouds, data mining, text mining, analytics, security and busi- capacity); platform as a service (computing capacity plus
ness continuity workloads offer higher potential ROI and middleware); or software as a service (computing capacity
lower risk. Test and developer environments are considered plus middleware plus applications).
good workload candidates for both public and private cloud. ● How will you create and deliver the services? This
includes the tools, procedures and governance required to
4. Determine the value plan, define, catalog, configure, deliver, monitor, measure,
Examine the ROI that cloud computing can bring, including bill and report on cloud-delivered services.
time required for initial payback as well as projected ROI. In ● How will users access the services? A service catalog that
determining ROI, start by calculating the absolute savings that enables end users to select, order and configure cloud-
will be realized from all facets of IT operations related to the delivered services is an essential component of a cloud archi-
workload in question—including, for example, hardware costs, tecture. In addition, for private clouds, the architecture
software licenses and upgrades, system administration, system should address providing an operational console for service
support, end user support and provisioning. ROI calculations delivery and operational managers.
can also include business-related measures, such as increases in
user productivity and resource utilization; avoidance of capital Underlying all three domains is the cloud platform—the
expense; and reduction of risk due to higher availability. architecture layer that defines the dynamic infrastructure for
cloud delivery, as well as the service management disciplines
In addition to looking at first-year savings, IBM recommends that enable service delivery and end user access.
that clients project operating costs of both the legacy IT envi-
ronment and the cloud environment over three years to calcu-
late ROI over a longer term. Our research has shown that
while all IT costs may increase over time, typically cloud
computing costs will increase at a lower rate—resulting in
increasing value.


6. Implement the IT strategy, roadmap and cloud services
A key component of an IT strategy leading to cloud is condi- The IBM cloud experience
tioning your infrastructure for cloud delivery. This may
In addition to numerous customer engagements, IBM has
include virtualizing and automating existing systems, adding
put cloud computing into practice in its own data centers
the service management capabilities requisite for cloud
around the world, and is able to draw on best practices
computing.
to share in customer implementations. Here are some
examples:
The roadmap for cloud implementation defines a pilot project
or projects that will enable the IT organization to gain experi-
● Blue Insight. An internal IBM cloud computing imple-
mentation that provides information on demand and
ence with cloud computing delivery while end users gain expe-
analytics as a service to IBM employees to help them
rience with standardized applications and services delivered
make better business decisions.
over the network. This pilot could be delivered from a private ● Cloud Sandbox. This internal version of IBM Smart
cloud platform, or it could be a pilot using trusted public Business Development and Test on the IBM Cloud
cloud services. Either way, the pilot should deliver an isolated, (available in 106 countries as a free beta service at time
low-risk workload. of publication) provides users with self-service access to
virtual computing environments as well as environments
Building on lessons learned, the next steps in the roadmap will preconfigured with IBM developer and database
be to enable additional workloads and extend cloud services to software.
new users. Once fully implemented, the cloud environment ● IBM Research Compute Cloud (RC2). RC2 is a self-
will meet end-user needs through a single portal with a cata- service, on-demand IT delivery solution that enables
IBM researchers to quickly and easily configure comput-
log of services that are automatically provisioned through the
ing environments needed for research projects.
cloud.
● IBM Learning Centers, Europe. The internal cloud com-
puting solution for IBM Learning Centers in Europe
The Cloud Computing Adoption Framework, developed by improved infrastructure utilization by 80 percent and
IBM, can be a helpful tool in following these steps to imple- reduced staff by 30 percent.
mentation. The framework establishes common definitions for ● IBM Computing on Demand. This service provides an
cloud computing delivery models and services, illustrates the enterprise-class public cloud computing solution that
key capabilities to consider in developing cloud computing enables clients to tap into IBM computing resources by
strategies, and identifies key aspects required to successfully the hour, week or year.
execute that strategy.

For more information
To learn more about cloud computing at IBM, please visit the
following Web site: ibm.com/cloud

Additionally, IBM Global Financing can tailor financing solu-
tions to your specific IT needs. For more information on great © Copyright IBM Corporation 2010

rates, flexible payment plans and loans, and asset buyback and IBM Corporation
disposal, visit: ibm.com/financing New Orchard Road
Armonk, NY 10504
U.S.A.

Produced in the United States of America
January 2010
All Rights Reserved

IBM, the IBM logo and ibm.com are trademarks or registered trademarks
of International Business Machines Corporation in the United States,
other countries, or both. If these and other IBM trademarked terms are
marked on their first occurrence in this information with a trademark
symbol (® or ™), these symbols indicate U.S. registered or common law
trademarks owned by IBM at the time this information was published.
Such trademarks may also be registered or common law trademarks
in other countries. A current list of IBM trademarks is available on the
Web at “Copyright and trademark information” at ibm.com/legal/
copytrade.shtml.

Other product, company or service names may be trademarks or service
marks of others.

References in this publication to IBM products or services do not imply
that IBM intends to make them available in all countries in which
IBM operates.
1
IDC eXchange, IDC’s New IT Cloud Services Forecast: 2009-2013,
p=543, Oct 5, 2009.

Please Recycle

CIW03062-USEN-01

Dispelling The Vapor Around Cloud Computing –Network Management

More Related Content

What's hot (15)

Similar to Dispelling The Vapor Around Cloud Computing –Network Management (20)

More from IBM India Smarter Computing (20)

Recently uploaded (20)

Dispelling The Vapor Around Cloud Computing –Network Management