Telco Cloud, how can it provide flexibility for traditional enterprise

Issue 4
Telco Cloud, how can it provide
flexibility for traditional
enterprise apps?
Telco cloud, how can it provide flexibility
for traditional enterprise apps?
2
Telco cloud, how can it provide
flexibility for traditional enterprise
apps?
8
From the Gartner Files: Exploring
Cloud Management Trends and the
Actions to Take
12
About Telefonica Business Solutions
Cloud; it’s in a TELCO’s DNA.
A common misconception is that the Cloud market is dominated
by Bulk Infrastructure Players such as Amazon or Microsoft, but not
everybody knows the TELCO’s capabilities in cloud. This is not the
case, however, as telecommunication providers have similar business
models, capabilities and infrastructure to enable them to offer an
empowering and competitive Cloud service. Cloud has become a
core part of a Telco’s business. Telcos historically have designed largescaled, shared infrastructures for both consumer and enterprises. It is
true that some changes have occurred since the first PSTN to now, and
that cloud services have been constantly evolving since the early years
of this 21st century. It is also true that Telcos have the demonstrated
experience of building and operating these large-scale missioncritical services, proven by services such as MPLS for Large Enterprises
in recent years.
One of the most important differences between Telcos and hyperscale
is the capacity to build up the various elements of infrastructure
within their own existing services. Telcos are used to partnering
and often business models have been based historically on buying
technology from a 3rd party while the hyperscalers decided to design
their own servers, storage and even network elements. Strategically,
however, when Telco’s develop their services, they do so within
common standards, for instance within the GSM Alliance which
enables seamless standards within which to partner and offer a more
comprehensive solution. This is the fundamental reason Telefonica
took this strategy, to enable open standard guarantees interoperate
with the Telco ecosystem. Not operating within these open standards,
hyperscalers are creating their own barriers which could limit their own
flexibility to their clients.
On the contrary, telcos have several advantages and several key
differentiators, such as: capillary strength (capillary infrastructure,
customer service and commercial power), along with data
sovereignty and trusted scalable performance which all continue to
make regional Telcos truly credible players in the Cloud market.
Telco Cloud, how can it provide flexibility for traditional enterprise apps? is published by Telefonica. Editorial supplied by Telefonica is independent of Gartner analysis. All Gartner research is © 2016 by Gartner,
Inc. All rights reserved. All Gartner materials are used with Gartner’s permission. The use or publication of Gartner research does not indicate Gartner’s endorsement of Telefonica’s products and/or strategies.
Reproduction or distribution of this publication in any form without prior written permission is forbidden. The information contained herein has been obtained from sources believed to be reliable. Gartner disclaims
all warranties as to the accuracy, completeness or adequacy of such information. Gartner shall have no liability for errors, omissions or inadequacies in the information contained herein or for interpretations
thereof. The opinions expressed herein are subject to change without notice. Although Gartner research may include a discussion of related legal issues, Gartner does not provide legal advice or services and its
research should not be construed or used as such. Gartner is a public company, and its shareholders may include firms and funds that have financial interests in entities covered in Gartner research. Gartner’s
Board of Directors may include senior managers of these firms or funds. Gartner research is produced independently by its research organization without input or influence from these firms, funds or their
managers. For further information on the independence and integrity of Gartner research, see “Guiding Principles on Independence and Objectivity” on its website.
2
Telcos have expanded their services within a very regulated
market (especially within EU), this has meant the investment and
strategy decisions have been made very close to local jurisdictions.
Hyperscalers, however, have been working in a globalized context,
following the dotcom boom, but may find it more difficult to be able to
offer the localized regulation compliance Telcos can facilitate.
What were the main advantages of this virtualization technology?
The essence for all players in the market, however, remains the same
for both sides of this argument. The final service is irrelevant: whether
voice, data, computing, storage,… the really important thing is the user
experience, the real time response and the flexibility to fit quickly to the
customer’s requirements and market needs.
• Dramatically lowers costs to redirect investment into value-add
opportunities
From virtualization to cloud.
History is once again repeating itself. Just like with the data center
vitualization wave, large enterprises continue to adopt, adapt and alter
requirements for their Cloud services. The first clients started to move,
test and develop their environments towards virtual environments.
Later, basic productive services, such as directory, file and print
services were then also added to the virtualized environment. Today,
things are very different. Over the last couple of years, enterprises
wanted to engage in data center consolidation projects in order to
revamp their infrastructures and improve the TCO and agility that these
virtual environments provided. The virtualization technology maturity
was, at that moment, unable to support intensive network rates
and big computing VMs sizes. Later, after several software updates,
software vendors solved their technical limitations and in 2010 many
enterprises had their first fully virtualized data center. In parallel many IT
Outsourcers found within this virtualization technology a lever to speed
up the transition phases to take full control of a client’s IT departments.
• Reduces complexity to simplify operations and maintenance
• Improves high availability to minimize downtime and IT service
disruption
Customers then begin to demand other features such as flexibility and
pay-per-use, in turn, testing drivers to start to trial the Cloud services
with the same life cycle management for their IT service catalog
levering on the benefits of the Cloud.
TELCO cloud services; the lesson learns
Our Cloud portfolio has been evolving since 2007, from shared
platforms in virtual farms to managed infrastructure based in a public
cloud portal, and then finally on to a fully virtual data center to fulfil
with all requirements for large enterprises.
• Virtual Hosting, first step as an IaaS provider: IT & communications
from a single point of contact
• Cloud Portal—our first self-service portal automating computing
and storage layers
• Virtual Data Center —new philosophy from a service provider
perspective
FIGURE 1
New features in cloud services thanks to technological advances.
Our Cloud Portfolio Has Been Evolving Since 2007…
Cloud Automation
Full Datacenter
& Services
Automation
Infrastructure
Automation &
Orchestration
Shared
Platforms
2007
2011
2013
Source: Telefonica
3
FIGURE 2
How Do We Evolve our Cloud Service?
Let’s Listen to the Enterprise Requirements…
Critical
Applications
performance on
shared platforms
SLAs and
Business
Applications
Monitoring
Recovery and
data protection
policies
Legacy
Systems
Integration
Nowadays,
Service providers
don’t fulfil all these
key functionalities for
Enterprise customers
Disaster
recovery and
Business
Continuity Plans
Maximize
expenditure
Agile and
flexible
management of
infrastructures
Security and
Regulatory
Compliance
Progressive
transformation
of
environments
to Cloud
Platforms
Benefit from
using a mix
between
private and
public
architectures
Integration
within Business
Processes
Source:
Telefonica
DISCOVER, DISRUPT, DELIVER
How have Telefonica evolved our Cloud service? Fundamentally, we
have been listening to our clients in order to fit our services with their
enterprise customer requirements. We are using LEAN technique in
the early stages of our developments, working with our customers
and learning quickly to adapt main features directly in line with their
demands.
Enterprise customers and the Cloud.
What have we learned during this journey?
Therefore, any new enterprises are likely to adopt the Cloud paradigm
as the core of the design of their IT strategy. This is certainly true for startups with a strong technological base who can develop their business
around the Cloud. Only Cloud can offer them the scalability needed in
order to achieve the demanding growth figures necessary for them to
survive.
Cloud in enterprise is here to stay, but adaptation and flexibility is
required.
While VMs lineally grow, vCPUs double their growth per year and
vRAM is exponentially increasing. More powerful VMs are required
due to more critical applications being virtualized in the Cloud.
The next wave; UNICA platform
Combining together core and traditional communications services
with cloud platforms is a huge challenge for Telcos, yet is one
which presents a unique opportunity to directly compete with the
hyperscalers. If Telcos were able to put all internal IT, core and auxiliary
core network services and cloud services together, with the same
underlying technology they are likely to be able to aggregate the same
economy scales whilst reducing drastically the price per GB of storage
or computing Ghz.
4
Nobody doubts that the cloud paradigm brings enormous advantages
to the management of IT for enterprises and that it fosters the flexibility
and growth of the IT systems. It moves the center of gravity of the
enterprise IT away from the management of infrastructure to the delivery
of value in terms of applications.
However, for a more traditional enterprise, the adoption of Cloud is not
so simple- for a number of reasons. The pace at which a traditional
enterprise can adopt an emerging technology or paradigm is clearly
slower, yet also their requirements are fundamentally different to those
of a newly-born organization.
A service provider targeting the corporate market with its Cloud services
has to design the offering to fully fulfill the requirements and needs their
customers in order to be successful. At present, well-known hyperscalers
are beginning to reach a size that makes them very competitive in terms
of price, as well as evolving their offerings to be able to incorporate more
and more functionalities at a speed never seen before in the industry.
FIGURE 3
What have we learned in this journey?
Cloud in enterprise is a fact, but adaptation is required
VMs
vCPUs
vRAM (GB)
≈ x9
≈ x6
2009
2010
2011
2012
While VMs lineally grow….
vCPUs double their growth per year
… and vRAM is exponentially increasing
More powerful VMs are required
Critical applications are being virtualized
on the Cloud
Source:
Telefonica
DISCOVER, DISRUPT, DELIVER
Other service providers are being forced to look at the wishlist of the
corporate IT organization to be able to compete successfully with the
other players in the market.
IT in a corporate environment is incredibly complex. The organizations
have been creating solutions and products over a very long period
of time, with a great number of different technologies, providers,
software and hardware. Quite often, each system was designed to
solve a specific business requirement, and architected to provide
the maximum efficiency. In many organizations the IT has been
fragmented: different units and groups have been building their own
solutions and therefore created a number of siloes that pose a real
challenge to the IT department in order to keep them up and running.
This has been made worse by the increasing numbers of mergers and
acquisitions in recent years in which IT departments have to maintain
very different solutions working together in what are often unstable
situations.
Therefore, when CIOs evaluate Cloud Solutions they frequently find
that a pure standardized solution does not fit well for their business.
In an ideal world, using a standardized “off-the-shelf” solution would
be the option of choice. In the real world, however, a degree of
personalization is desired to be able to tailor fit their requirements.
An enterprise Service Provider, therefore, needs to recognize these
individual needs and then be able to implement the complete solution
with minimum disruption to their customer.
Corporate customers have typically been operating their IT in Data
Centers (owned or leased) and have a deep knowledge of the
technologies and capabilities around them (computing, networking,
security). They are able to build complex solutions, monitor and
operate them. They would like to maintain the same level of control
over the Cloud infrastructure as they have now on their physical
one. In fact, moving to many Cloud solutions represents a loss of
functionality and control that are not willing to accept. Or that moving
to a native Cloud environment creates the need of costly and painful
migration of applications that could not be carried out. Consequently,
solutions that provide an equivalent level of control to the ones they
are used to will have a real appeal for customers.
One of the key aspects of cloud is the capability of self-management.
Typically carried out through web portals and APIs, it empowers the
IT department to increase their productivity, substituting long and
tedious processes and allow different groups within the organization
and to interact directly with the portals. Corporate Customers are
also adopting Managed Service Models, in which they transfer the
management function to the service provider. A combination of these
two models, with the customer being able to control and supervise
the elements managed by the Service Provider or a third party, is now
gaining traction in the market. It offers clear advantages over the
traditional outsourcing model in which the information provided to the
customer is limited and not in real time.
5
Another key customer requirement is how to manage their non-cloud
world. Enterprises have invested huge sums of money in hardware
and software that cannot simply be forgotten. It is clear that, for
a considerable time still, new systems and old ones will have to
coexist and work together. For some enterprises, migrating certain
aspects of their business to the Cloud represents a real challenge.
As a consequence, offerings that make legacy and new cloud
environments work together in a seamless way will continue to receive
a good market acceptance.
Service Providers that are able to offer a wide portfolio, from traditional
Data Center services (housing, co-location) to the newest cloud
solutions and, more importantly, orchestrate these elements to make
work them together as one environment, will be able to better serve
their customers.
The same old story
The differences between a Corporate’s market and the Cloud native
market will eventually blur in some years. The adoption of Cloud
will follow a similar track to that of Virtualization technologies some
years ago. First, non-critical or new developments were deployed in
Cloud environments, then gradually core systems are also eventually
migrated to the Cloud. Some legacy systems may survive but it will only
be within some niche applications and platforms. Whatever happens,
the core element of the value proposition of a Service Provider must be
the freedom for the customer to choose their desired elements. Service
Providers must partner with the customer in the process of evolution of
the Corporate IT to be a trusted advisor and be a creator of value and
not a consumer of resources.
Born in the cloud vs traditional IT.
The latest report from Gartner, which follows later in this Newsletter,
“Gartner Exploring Cloud Management Trends and the Actions to Take”
remarks where the power of cloud computing remains: “enterprises
have realized that the power of coud computing is in the speed and
agility gained in developing and operating cloud-based applications”,
It is true that elasticity is a key feature of Cloud, and it is key, because a
flexible infrastructure that can grow or shrink as business needs change,
is really attractive to an enterprise wanting to become more efficient.
Nevertheless, as stated before, enterprises host a range of legacy
applications which were not developed within Cloud and, therefore, the
full versatility and power that Cloud can contribute to their business is
being underused with their older applications unable to exploit them.
The report also comments that companies are; “refactoring existing
applications to take advantage of the elasticity of the Cloud”, however
in this transition period, until legacy application can be transformed,
is necessary to use a framework which allows the coexistence of both
types of applications. This is where the Hybrid Cloud plays a key role in
fostering the harmony of both, since it solves the existing gap between
legacy applications and new apps providing a common management
interface. It also remarks the need to “Ruthlessly standardize (for both
private and public cloud computing) to enable efficient reuse and
automation of the infrastructure and application services”.
With both of these concepts in mind, Telefonica’s Hybrid Cloud has
been designed to follow this criteria: using both dedicated and shared
infrastructure (our shared Cloud product is called Virtual Data Center).
But how do you merge both Virtual Data Center and Private Cloud
successfully?
“Everything in any cloud”: Their Hybrid Cloud is Telefonica’s offering to
their customers. Through this Hybrid Cloud, Virtual Data Center can
be connected to a Customer’s Private Cloud:
• Single point of management
• Option to balance loads between one and another
• Everything has the most end user-friendly interface
Telefonica Hybrid Cloud is a powerful, scalable option especially
for customers doing business within regulatory limits, dealing with
licensing issues or sensitive data. Telefonica Hybrid Cloud allows
customers to transfer and maximize the Cloud benefits for multiple
forms of applications whether legacy or “born in cloud”, adapting to
the specific needs of each customer.
6
For example:
• Can I develop a new application for
my business directly in my Cloud
environment? Yes, the customer can test
their application in the Virtual Data Center
and make use of Cloud work-load features
for their production purposes. This is
frequently used when developer agility is
critical, they must be aware, however, of
compliance and security concerns related to
the production data.
• Does a new application “born in cloud”
need data from my Legacy Application?
The solution uses Split-tier architectures
to accommodate the new application
in VDC and connects via high-speed
communication networks to the dedicated
environment in which the legacy application
sits. Telefonica network capillarity makes this
possible as a real, workable option since
latency issues are minimized.
FIGURE 4
Hybrid cloud will evolve to Multicloud: Manage expanding portfolios of public
and private cloud platforms _
NewApps “Born In Cloud”
Productivity And Cloud
Infrastructure Control
Policy-Based Orchestration
Governance
Operational
Control
Private
Cloud
• What about applications with unpredictable demands? For
new applications with unpredictable demands, the flexibility of the
shared environments work incredibly well. Stable applications, with
steady demands stay on the private Cloud. Once the unpredictable
application reaches a steady state, it will be moved to private Cloud.
All applications run on the same orchestration framework allowing
the known advantages of operational efficiency.
• What happens if my applications or Cloud solution fails? If the
production environment resides on a private Cloud, our solutions
offer a disaster recovery service in our Virtual Data Center, ensuring
the consistency of the data. If an event of a disaster occurring, the
service includes an automated mechanism to activate the recovery
service.
By combining both models, Telefonica offers Hybrid Cloud solutions
that allow customers to build their own Private Cloud for critical
environments or those containing sensitive data, while also using the
Public Cloud for their additional consumption needs, peaks in demand,
back-up environments, and much more.
Cost
Hybrid
Legacy Applications, Core Apps
• Where are the application users? If they are geographically
dispersed, a good practice would be to develop on the existing
vacancy infrastructure and use public cloud in production time to
provide scalability and/or geographical coverage. This applies
when there is excess private cloud capacity capable of developing,
but production applications need higher scalability or better multigeographic capillarity.
“Multicloud will help
companies balance demands of
cloud developers and
technology managers”
• 
• 
• 
• 
Bursting
Load Balancing
Contingency
Additional Capacity
VDC
Source: Telefonica
At Telefonica, we predict that Hybrid Cloud will evolve from hybrid
homogenous Cloud environments to orchestrate heterogeneous
Cloud environments, not only based in our own Cloud options but
also with other Cloud providers. The next generation Cloud services
will include Cloud brokering facilities. We believe they are the ideal
transition from the traditional delivery model to a new model where IT
infrastructure and commodity IT services are sourced from a selection
of best-in-class service providers.
These new components act as an over-lying layer, capable of
providing an overview of how systems are supporting the business
and providing consistency in terms of access. It offers customers
the transparency to understand where the systems are allocated
as this new layer manages the relationships between Cloud service
providers and consumers. The solution covers different aspects,
from Cloud providers governance processes, cost management and
monitoring.
The diagram above illustrates these new models:
One of the main concerns of CIOs is how much money is their
business is spending on IT that is not being accounted on their IT
budget. This new Cloud model also provides a valuable way of
managing this ‘shadow’ IT as well as limiting the inherent risks within
these practices such as information leakage or lack of control.
Source: Telefonica
7
From the Gartner Files:
Exploring Cloud Management Trends
and the Actions to Take
Next-generation hybrid cloud services will be
designed by I&O architects with developers
and brokering in mind to support faster time
to innovation. With clearer requirements,
hybrid cloud services will increase in success.
The cloud management market, however, will
continue to fragment and evolve.
Key Findings
• Changes in cloud management platform
(CMP) technology are being driven by
enterprise interest in platform as a service
(PaaS) and containers, as well as changes
in software development for cloud
applications.
• IT organizations desire multicloud service
management and governance, which
often limits the use of native public
cloud services to the lowest common
denominator, but their line-of-business
customers want to take advantage of all
cloud-native capabilities.
• IT organizations want to preserve flexibility
in using multiple cloud providers, and
desire to reduce lock-in.
• Increasingly, enterprises are seeking to
optimize the placement of IT services
across public cloud and on-premises
private clouds.
Recommendations
• Design your next-generation private, public
and hybrid cloud computing architectures
around self-service developer and
operational requirements.
• Bring together a multidisciplinary team that
fulfills the cloud service brokerage (CSB) role.
• Verify that vendor road maps address
infrastructure integration and policy
enforcement if you are looking at PaaS
solutions.
• Ensure that your selected CMP vendor has
a clear approach that is consistent with your
organization’s business strategy.
• Expect not to gain a lot of value in private
cloud computing for nonelastic applications.
8
Analysis
Recommendations:
Enterprises Focus on Cloud Services
for Rapid Release of New Business
Capability
• Expect not to gain a lot of value in private
cloud computing for existing nonelastic
applications.
Since 2007, enterprises have embarked on
private cloud computing initiatives that have
mostly failed due to complexity, lack of skills
and difficulty defining new processes that
cross technology domains. Even more so,
there were expectations that private clouds
could improve non-cloud-based applications
or magically solve problems resulting from
lack of standards and reusability. Now,
however, most enterprises have realized
that the power of cloud computing is in the
speed and agility gained in developing and
operating cloud-based applications. Many
have begun to implement agile methodologies
and continuous integration for their newly
developed Web, mobile and analytics
applications (what Gartner terms “Mode 2”).
They have also started to realize that agility
requires standardization of the infrastructure
and the software stacks, as well as completely
new processes that cross technology
domains (compute, storage, network, security,
middleware). This enables developers to get
what they need from IT through self-service,
eliminating tickets as a means of handoff
between groups in the organization for
provisioning infrastructure.
Enterprises are creating cloud applications
from scratch, and refactoring existing
applications to take advantage of the
elasticity of the cloud. The result is that most
implementations of private cloud computing
are following those of the public cloud – to
develop new applications that are dynamically
composed from standardized components
and bound to the infrastructure at runtime.
As a result, the next generation of private and
hybrid cloud computing solutions will bettersupport agile methodologies and DevOps
initiatives, driving the necessary improvements
in time to market that Web, mobile and digital
business applications demand. In addition,
these private/hybrid cloud computing efforts
will be more successful than with past efforts,
as a result of having the right users and targets
in mind when developing the end-to-end
implementation. They will also be able to more
easily take advantage of the elasticity of hybrid
cloud deployments.
• Design your next-generation private,
public and hybrid cloud computing
implementations around self-service
developer requirements to speed time to
market of new applications.
• Design your applications to be multicloud,
in that they can be deployed in the public
or private cloud, or span both for a hybrid
computing service.
• Ruthlessly standardize (for both private
and public cloud computing) to enable
efficient reuse and automation of the
infrastructure and application services.
• Design your new applications with life
cycle in mind by enforcing standard
components and operational automation.
• Include both complete redeployment, as
well as incremental updates to support
business availability requirements,
when designing your production release
management processes. While dynamic
applications improve time to market and
empower developers through self-service,
not all applications entering production
will be completely dynamic.
• Look for cloud management platform
technologies that extend beyond
provisioning and deprovisioning
infrastructure to application stacks,
environments and application release
automation capabilities.
A Focus on Developers Will Bring
CMP Features to PaaS Platforms
Most enterprises have implemented cloud
computing using traditional development
practices, including automated deployment of
traditional middleware and DBMS (also called
infrastructure as a service plus middleware
[IaaS+]). Increasing focus on developeroriented cloud services and time to market
will increase interest in PaaS – for both public
cloud and private cloud implementations.
9
While there are advantages, PaaS platforms
have more lock-in associated to them than
IaaS, that is, they are harder to migrate to
another platform or provider.
Today, most PaaS platforms have little
visibility into and control of the underlying
IT infrastructure and data center locations.
Instead, the focus has been to enable
developer productivity, but then deploy locally
or require separate (and often manual)
processes to deploy the applications to the
right places and to enforce infrastructure
policies. This will change in the nextgeneration PaaS platforms as they become
more infrastructure-aware, cloud-serviceprovider-aware and intelligent. More
CMP-like features will be embedded in PaaS
platforms, for example, to blueprint the
logical model of an application and to design
deployment, elasticity and management into
the application as it is built. Infrastructure,
therefore, is treated much more like
middleware in the PaaS platform. This will
result in PaaS platforms integrating with
CMPs or resource management layers, such
as OpenStack and cloud service provider
platforms.
Making the PaaS platform infrastructureaware does not mean that developers
need to focus on the infrastructure. Rather,
it means that infrastructure architects can
develop policies on data center (or external
service) placement, availability, continuity and
recovery requirements, as well as standards
for reusability (to reduce diversity and ongoing
management costs). These policies can be
enforced based on various factors, like the
specific user, the type of application being
built, the stage of the development process
and the mission criticality, for example. As a
result, infrastructure and management can be
better-automated and -optimized throughout
the application life cycle across both public
and private cloud services.
Embedding or integrating CMPs (and IaaS)
into PaaS will cause yet another evolution
of CMP requirements and functionality. It
also means increased competition between
CMPs and PaaS vendors. It will also further
fragment the market (see the CMP Vendors
Will Evolve and Consolidate Over the Next
Two to Five Years section). Enterprises that
use public PaaS will likely have very little
visibility or control over the data centers and
infrastructure, while those deploying PaaS
internally will have the opportunity to apply
more uniform and consistent infrastructure
policies. It will be critical that enterprises
that desire data center and infrastructure
programmability and policies choose their
architectures (and vendor tools/road maps)
carefully to be sure that their PaaSs and
underlying management technologies
achieve their objectives with minimal
technical debt and sprawl. Moreover, new
management capabilities will be needed to
orchestrate across PaaS and IaaS services,
providing visibility and cross-cloud-provider
service management functionality.
Recommendations:
• For enterprises continuing with traditional
development practices, look to CMPs
to enable infrastructure and operations
policy enforcement for private, public and
hybrid cloud services.
• For enterprises looking at public and
private on-premises PaaS platforms,
evaluate road maps, ensuring they
include data center and infrastructure
visibility and policy enforcement. Note
that public PaaS typically will not offer
the depth of visibility and control of the
infrastructure that private PaaS options
afford. For enterprises pursuing both
traditional and PaaS development,
evaluate and select CMPs and
PaaS platforms that enable multiple
development methodologies, but with a
common policy framework.
• As with any cloud strategy, assess
processes that would be used to migrate
from one provider to another early in
the process, in order to reduce risk and
vendor lock-in.
Acceptance of Public Cloud Drives
Greater Investment in Cloud
Strategy, Management and
Brokering
Many IT organizations were caught off guard
by the amount of public cloud services their
business units acquired with little to no IT
oversight or any particular cloud strategy in
place. Now that public cloud IaaS services
are both accepted and welcomed at most
large enterprises, it has become apparent
that there is a need for policies, standards,
compliance, life cycle management and
ongoing financial management. Often, as
a result of compliance, risk or cost factors,
enterprises are beginning to refocus on
architecture standards to drive more value
and speed out of cloud computing while
reducing risk and cost. Moreover, they are
taking their experiences gained with cloud
computing and putting together a strategy
that makes sense for their enterprise. This
strategy should include defining the optimum
mix of cloud services that should be run
in the public versus private cloud (versus
9
traditional or virtualized infrastructure);
determining what factors should drive public
versus private cloud adoption; and driving
architectural design standards that foster
reuse and embedded manageability.
As a result, a greater number of enterprises
are creating the role of cloud service broker,
who, working with the cloud architecture
group, develops strategy around what should
run in the public cloud versus on-premises,
determines what applications should migrate
to SaaS, and selects the cloud providers that
meet enterprise requirements. This role is
multidimensional and often includes skills of
business management (business process
owners), service management, application
management, data center/infrastructure
management, governance, provider
management and financial management.
With so many new cloud services frequently
coming into the market, the CSB also works
with architecture groups/CTO in assessing
whether these offerings provide value to the
enterprise.
Increasingly, CMP technology will add more
brokering and assessment capabilities to
support the cloud service broker role. This
includes, but is not limited to, marketplace/
aggregation functionality, workload/SLA
assessments across cloud providers, cost
optimization, performance optimization,
orchestration and billing/chargeback across
providers. Some organizations will use CMP
technologies for API translation layers and/or
migration across cloud providers, especially
those using traditional development
techniques. Others will use PaaS platforms
to do the translation, or alternative means of
enabling migration between cloud providers.
While some vendors tout the need for
dynamic movement between cloud providers,
most enterprises do not require such
capability today. Rather, they require bursting
for additional capacity and the ability to move
between providers when contracts are “up,”
if they are unhappy with their service provider
or if their cost profile changes.
10
Recommendations:
• Develop a strategy for cloud-native
and -optimized applications, including
embedding automated management
processes. Then formalize your cloud
strategy if you have not already done so.
• Implement a cross-functional role of cloud
service broker to aid in determining the
optimum mix of services that should run
on-premises versus off, and in brokering
cloud services for the enterprise.
• Look for CSB-enabling functionality that
can be used for service aggregation
(marketplaces), workload “fit” to cloud
providers, cost comparisons, cloud
service optimization and multicloud
billing/chargeback, when assessing CMP
technology.
• Build an exit strategy for every cloud
provider (or a holistic methodology for
all). Make sure that you can move your
services in a reasonable time should
the provider stop performing, if you are
dissatisfied or if you change your strategy.
CMP Vendors Will Evolve and
Consolidate Over the Next Two to
Five Years
The cloud management platform market
has existed for over five years. Gartner tracks
over forty CMP solutions. These solutions
can be grouped within the following vendor
categories: infrastructure software, large
IT operations management, open source,
integrated infrastructure systems, cloud
service brokering vendors adding CMP
capabilities, and emerging vendors. No
single product is (or set of products are)
significantly distinguished from others in the
market. While there have been successful
CMP implementations, many products lack
features and/or have shown to be very
difficult to deploy.
The CMP industry is also being challenged by
evolving adjacent products and technologies.
On the product side, products that originally
focused on the CSB-enabling area (for
example, those from RightScale and CliQr
Technologies) are increasingly adding
CMP features. Combining CSB and CMP
capabilities will be highly desirable over the
next few years, because we anticipate that
enterprises that have deployed or plan to
deploy private clouds will increasingly need
to integrate public cloud services.
We believe that most CMPs that developed
primarily to manage on-premises and hybrid
cloud services will increasingly add brokering
functionality (such as cost comparisons, bestfit analysis, orchestration across multicloud
environments and longer-term migration
between providers) to their products. Similarly,
PaaS offerings that have initially only focused
on application provisioning are beginning to
add infrastructure provisioning services (as
noted in the A Focus on Developers Will Bring
CMP Features to PaaS Platforms section).
Finally, IT process automation (ITPA) tools
(in combination with a self-service portal or
service catalog) are also encroaching in the
CMP area. These tools allow orchestration
of a limited set of IT resources without the
complexity and “heaviness” of many CMPs.
On the technology front, containers
(because they augment, or even supplant,
virtual machines as the basic unit of
processing within a cloud environment)
have the potential to be very disruptive to
the CMP market. In this case, container
tools for orchestration, such as Kubernetes
from Google, could evolve as a possible
replacement for CMP orchestration
tools, thus requiring CMPs to evolve their
approach yet again to governance and
multicloud management. The main point is
that, as applications get rearchitected with
cloud-native and -optimized paradigms,
fragmentation in how the application models
are created will create further segmentation
and evolution in the CMP market.
11
These dynamics will force the market to
consolidate. While there has already been
some consolidation within the market (for
example, HP acquiring Eucalyptus), more
consolidation is anticipated.
• Objectively evaluate your readiness for
deploying a cloud management platform.
If your organization is not ready to deploy,
a first step could be increasing your
level of virtualization while adding some
automation capabilities.
Recommendations:
• Recognize the volatility and evolution of the
CMP market, and proceed cautiously and
tactically. Choose vendors that you believe
will evolve with your needs, and prepare
yourself to switch if they don’t meet
your needs or if your business strategy
changes.
• If you are ready to deploy a CMP, factor in
the following as a part of your selection:
strategic vendor relationships, existing IT
operational management tools, application
development (AD) strategies (this includes
what your AD team’s plans are for PaaS
and containers). Ensure that your selected
vendor has a clear strategy that is
consistent with your organization’s strategy.
• Use an orchestrator that has a lot of
out-of-the-box adapters for public and
private clouds, and also the ability to
create adapters quickly for unsupported
environments. In addition, ensure that
the scripts and workflows developed are
platform-agnostic to enable portability.
• Define management requirements for
public and private cloud services, and
evaluate the need to either adopt a CSB,
or leverage brokerage tools so that IT
becomes the broker to the organization.
Source: Gartner Research, G00236106, Donna Scott,
Milind Govekar, Dennis Smith, 11 February 2015
11
About Telefonica Business Solutions
Telefonica Business Solutions, a leading provider of a wide range of integrated
communication solutions for the B2B market, manages globally the Enterprise (Large
Enterprise and SME), MNC (Multinational Corporations), Wholesale (fixed and mobile
carriers, ISPs and content providers) and Roaming businesses within the Telefonica
Group. Business Solutions develops an integrated, innovative and competitive portfolio
for the B2B segment including digital solutions (m2m, Cloud, Security, e-Health or
Digital Marketing) and telecommunication services (international voice, IP, bandwidth
capacity, satellite services, mobility, integrated fixed, mobile, IT services and global
solutions). Telefonica Business Solutions is a multicultural organization, working in over
40 countries and with service reach in over 170 countries.
https://twitter.com/TelefonicaB2B
12