Understanding IBM's Hybrid and Multicloud Strategy
Understanding IBM's Hybrid and Multicloud Strategy
Judith Hurwitz
President & CEO
Daniel Kirsch
Managing Director
Sponsored by IBM
A Hurwitz wHite pAper
Introduction
We recently had the opportunity to speak to a CIO who helped his company
make the strategic transformation from an inflexible and chaotic infrastructure
to a hybrid computing approach. The business was changing too fast to simply
carry on with business as usual. The company had to compete with emerging
competitors that did not have to contend with any technical or business legacy.
The CIO’s mandate was to create an IT environment that was innovative, flexible,
scalable, manageable, and safe.
After many years spent focusing on reducing capital expenses through the
introduction of public clouds it became clear that this approach was not able to
meet the demands of the enterprise. The only viable solution was to embrace
open source standards implemented in a hybrid and multicloud environment.
At the same time, his organization was able to extract important intellectual
property and best practices trapped within existing applications that had to
be transformed into modular services. To gain acceptance from management,
the new cloud environment had to have security and compliance at its core.
Administrators also needed to have consistent ways to manage integration and
data services. The predictability and manageability of the hybrid cloud now
enables the business to move fast to create new value in light of competitive
threats, while at the same time delighting customers with innovation and new
products. All of this had to be accomplished while keeping costs under control.
of deployment models that best suited their objectives. He moved forward with
platform services that would support a continuous software development and
deployment model combined with ubiquitous and security services. At the same
time, he began the process of modernizing and refactoring existing applications
so that they were modular and flexible. He had to do all of this while maintaining
customer satisfaction.
It isn’t enough to have a
It is clear from the experience of this CIO and many of his peers that business as world-shattering idea if you
usual is dead. Instead, enterprise leaders must be on a quest to innovate quickly don’t have the technology
while maintaining the advantages of their deep intellectual property, their infrastructure to support
installed base and revenue. It is not a simple transition. Successful enterprise change and execute at scale.
leaders understand the challenge and have selected a hybrid cloud environment.
One size does not fit all. Rather, to be successful, businesses are focused on a
multicloud, open source strategy that allows them to innovate, scale and protect
their assets and their customers.
It isn’t enough to have a world-shattering idea if you don’t have the technology
infrastructure to support change and execute at scale. We are entering an
interesting stage of the digital transformation movement. Established businesses
have sounded the alarm as they have watched startups – that have no legacy
systems – move fast with great ideas and new business models. These emerging
companies have indeed caused revolutions in many industries. Market
incumbents have been forced to rethink how they conduct business and how
they use technology. Although it might seem like these market stalwarts have no
chance against nimble, born-on-the-web competitors, established companies
have the benefit of decades of industry and customer data as well as intellectual
property and complex business rules. Additionally, these businesses know how
to scale to support a global footprint.
IBM’s cloud strategy has evolved as customer expectations for the cloud
have exploded. IBM has poured a significant investment into a hybrid and
multicloud foundation based on industry standards and open source. At the
core of IBM’s cloud strategy is the understanding that customers need a unified
and streamlined way to create modular application services, to transform and
modernize legacy, and to manage all services, no matter where those workloads
reside. All of these services need to be cataloged, secured, managed, and
governed based on policy rules.
IBM private and public clouds as well as on any third-party public cloud and
services at the edge. The power and versatility of open source, including Linux,
open APIs, and Kubernetes, has helped IBM transform its software so that it has
become modular and has unified services that can be more easily managed.
The IBM customer base is complex and therefore there are different
requirements for the transition to the cloud. IBM has defined three entry points
or models to support customers.
Migrating existing services to the cloud. Most large enterprises must
contend with a large number of outdated legacy applications and
infrastructure. These businesses need a path to modernize and refactor their
applications so that they gain the flexibility and agility needed for business
transformation. These businesses may begin by lifting existing applications
so they can be hosted on a cloud through the use of virtualization through
VMware services. This entry point is designed to help businesses that want to
take expenses out of their infrastructure.
Modernizing existing applications and services. Organizations are
transforming their applications with a container platform in order to
modernize and migrate existing applications. The computing environment
is analyzed to determine which applications can be reconstructed as
microservices and which applications will need to be redesigned. Once
containerized, it is possible to link those microservices as part of a cloud
foundation through the use of open standards. Once these existing
applications are refactored, developers can move quickly to create new
applications.
New Cloud Native Innovative Applications. Using cloud native services
such as Knative and serverless, developers can create next generation
applications that can be deployed within an open source Kubernetes
environment.
One size does not fit all. Rather, to be successful, businesses must focus on a
hybrid, open source plan that allows them to scale and protect their customers.
It is necessary to leverage different computing models for different use cases.
There are complex business applications that cannot easily be rewritten
without a massive and expensive effort. Many of these services reside in the
data center or are hosted in an enterprise cloud. In addition, there is a need for
public cloud services to support agile DevOps requirements. Public clouds are
ubiquitous and will continue to grow and mature. There is not a single public
cloud that all customers demand. Rather, there are a variety of public clouds
that are adopted across lines of business. In addition, there are mission-critical
workloads containing strategic data that businesses do not want to place in
the public cloud. Considerations such as security and compliance and latency
are core to why businesses are insisting on a hybrid environment. Given these
considerations, IBM has adopted a hybrid and multicloud approach.
IBM has long been making significant investments in open source technologies,
-- for example contributions to Linux, Hadoop, Spark, Kubernetes, Calico, Helm,
Istio and Knative. This investment in open source has increased dramatically over
the past five years as IBM has begun to transform its public and private cloud
platforms based on open source standards.
One of the key rationales for IBM’s pending acquisition of Red Hat is the reality of
Red Hat’s business model and customer engagement. (The acquisition is subject
to regulatory reviews. Until close, the parties are separate and independent
companies.)
Red Hat’s product offerings are all based on commercial open source software,
ranging from the most popular Linux distribution, Red Hat Enterprise Linux
(RHEL), to other offerings like Red Hat Ansible, OpenShift, and CloudForms,
which provides a variety of automation and management technologies to
improve operational agility.
Both IBM and Red Hat have containerization at the core of their strategies. Both
companies focus on managing workloads across the enterprise and assume
a hybrid computing environment. Red Hat is well respected by the developer
community for its support of open source and its ability to deliver these
resources in an enterprise- safe manner. Additionally, Red Hat provides the most
important commercial Linux distribution that is the foundation for a majority of
enterprise customers. Both companies are focused on creating enterprise scale
solutions to support a large variety of solutions and tools all based on an open
source framework.
The benefits of Kubernetes to IBM cannot be overstated. IBM has made the
strategic decision to base both its public and private cloud on the same
Kubernetes orchestration. Kubernetes is a portable and extensible open source
platform for managing containerized workloads and services. One of the
values of Kubernetes for IBM is that it is optimized to support all infrastructure
platforms --including virtual machines-- to mainframes, and bare metal
infrastructure. Another key reason for IBM selecting Kubernetes is the expansive
ecosystem of tools and services that IBM leverages. This means that IBM is able
to leverage important Kubernetes infrastructure tools including Istio, Calico, and
Helm.
connection to use legacy services containerized on the IBM private cloud. For
the IBM Cloud Private (ICP), IBM has created a Kubernetes that has containerized
both IBM and OpenShift middleware. ICP also containerizes all services that are
native to the host environment including IDAP and management services in the
host.
One of the most important changes that IBM has containerized is its middleware
IBM’s approach to hybrid
integration services managed through Kubernetes. The common integration and multicloud security
platform includes existing IBM products such as API and APP Connect, MQ, is to create a set of built-
Aspera (secure high speed data transfer) and Event Streams. It has incorporated in capabilities that are
open source middleware offerings such as Jboss, TomCat, Kafka, and Liberty. All independent of the
of these offerings are provided in an “as a service” model on both the IBM public infrastructure layer.
and private cloud and on any third-party public cloud.
Visibility
One of the most difficult security tasks is to have consistent visibility into the
security of a hybrid and multicloud environment. IBM has developed what it
calls the Security Dashboard. This service provides visibility to the administrator
across all platforms. It provides access to logs from all of the services being
used through a series of APIs. It also provides a vulnerability advisor that helps
set policies for who is allowed to access a service based on compliance and
policies. The Security Dashboard (hosted in the IBM Cloud) can bring in data
from private as well as public clouds such as AWS and Azure. Third-party vendors
can integrate their services into the dashboard. The service keeps track of which
employees or contractors are permitted to access services.
Managing Access
One of the greatest security risks in a hybrid/multicloud environment is to
be able to control access at both the network level and at the individual Encryption is a key to any
level regarding who is allowed to access what service for what purpose. At data protection strategy
the network level, it is imperative to be able to identify malicious access and
– encrypting all data will
anomalous behavior. Through the access management services, customers can
gain insights from activity logs and aggregate those security events and logs in ensure that even if there
order to understand the risks across a multicloud environment. is a breach, the criminals
will not be able to read or
Protect Data and Applications understand that data.
Securing distributed data and applications is not easy. First, data may be at rest,
in transit, or in memory. No matter the form, all of the data must be secured.
The data may be managed within an object store, a cloud database, a virtual
machine, or in the data center. IBM is providing a set of hyper protect data
management services that is offered as a cloud service. This brings the highest
level of data security and provides ways to control that security. Encryption is a
key to any data protection strategy – encrypting all data will ensure that even if
there is a breach, the criminals will not be able to read or understand that data.
These security services can be implemented in a container, a virtual machine, or
on premises in a private cloud. IBM includes key management services, including
the ability for the client to be the only holder of the key – meaning that IBM itself
has no copy of the encryption key.
Customers that use LinuxONE or IBM Z as a cloud service are able to take
advantage of the platform’s pervasive encryption technology. Pervasive
encryption can encrypt data both at rest and in flight, and this type of
encryption doesn’t require application changes. This approach enables
companies to encrypt all their data by default with little compute overhead.
Because of the overhead of software encryption, businesses in the past have
been forced to choose which data to encrypt, leaving the remaining data at
risk. In nearly every online interaction, data is left unencrypted at some point
in the process. This point, when data is left unencrypted, gives wrongdoers the
opportunity to steal data. IBM Z and LinuxONE both have dedicated hardware
specifically tuned for encryption. The on-chip encryption co-processor is on
every compute chip next to the main processor and can encrypt up to 13
gigabytes (GB) of data per second per core.
One of the benefits of MCM is that it can scale out Kubernetes clusters and
dynamically add and remove nodes. As applications and services are deployed,
MCM provides standards-based Terraform as its load balancer. In fact, all of the
services that underlie MCM are based on accepted and emerging open source
standards. MCM is available either as a software license or as a managed service.
MCM also provides a consist way to manage integration. To reduce the footprint
of integration, the service provides a common user interface, common APIs,
and common API management, as well as common logging and management
services. The common APIs provide a platform for hybrid cloud data exchange
data and business process management services. One of the benefits of the
Integration service is that it permits businesses to incorporate existing enterprise
services such as billing and transaction management services so that they can
be seamlessly integrated with cloud services. MCM also plays an important
role in managing security. The only way to secure a distributed environment
is to containerize both workloads and data that provide consistent APIs. A
containerized approach enables applications, operations, and data services to all
The only practical way of understanding and managing this complex set of data
sources in a consistent and predictable way is to leverage advanced machine
learning algorithms. To address the challenges of managing, ingesting and
analyzing data from a variety of sources, IBM has built machine learning and
Artificial Intelligence capabilities throughout the platform. The machine learning
technology streamlines and automates time intensive tasks like data cleansing,
integration, and cataloging of business ready data.
IBM’s Cloud Private for Data (ICP for Data) is composed of a set of pre-
configured microservices that run on a multi-node IBM Cloud Private cluster.
The microservices enable the organization to connect to a variety of cloud and
on premises data sources including relational, NoSQL, IoT, and text data, so they
can be consistently catalogued, governed, transformed, and analyzed through
a single web application. This pre-integration means that the data layer doesn’t
require customers to integrate various data management and analytics offerings.
Many companies have tried to break down silos by copying data from different
operational systems into central data stores for analysis such as data marts, data
warehouses and data lakes. There are many problems that occur during this
data movement. For example, it is expensive, time consuming and cumbersome
to manage so many disparate data sources. Data can be inadvertently exposed
when data is disconnected from security and governance policies. To avoid the
challenges of replicating and moving data for analytics, IBM has developed a
data virtualization approach called IBM QueryPlex. Queryplex allows an analytics
application to submit a query that is then processed on the system where the In addition, all of the data
data resides. Results of the query are then sent back to the analytics application. services in the IBM Cloud
platform are delivered in
In addition, all of the data services in the IBM Cloud platform are delivered in containers and supported
containers and supported by Kubernetes-based orchestration services. The result by Kubernetes-based
of this containerized approach is that IBM hybrid database services can operate orchestration services. The
on any public or private cloud service including the IBM Cloud, Amazon AWS, result of this containerized
and Microsoft Azure. In addition, the platform has standardized APIs to support approach is that IBM
all services to ease the ability to integrate services. hybrid database services
can operate on any public
The IBM cloud data platform can support a variety of the most widely used
relational databases including Db2 and the Db2 Warehouse, Oracle, as well or private cloud service
as NoSQL, MongoDB, Postgres, and IBM’s Event Store for streaming data including the IBM Cloud,
from sources like IoT workloads. Within the next year IBM expects to bring on Amazon AWS, and Microsoft
additional data services platforms through its partnerships including Redis, an Azure.
in-memory datastore.
Premium Services
Once the foundational services are in place, the data platform is designed
to enable customers to add premium services – either those offered by IBM
or those offered by partners. These premium services are primarily used for
reporting, advanced analytics and artificial intelligence (AI). The Watson APIs
are available on the multicloud platform to add AI functionality to customer
applications. Some of these Watson APIs include Watson Assistant, to help build
chatbots; Speech to Text to convert audio and voice into written text; Tone
Analyzer to understand emotions and communication style in text; and, Visual
Recognition to tag, classify and train visual content. For customers who need
basic reporting and business analytics services, there are the Cognos tools and
dashboard. For advanced analytics such as predictive analytics and machine
learning, IBM offers tools and services including SPSS to build predictive
analytics solutions. In addition, IBM offers Watson Studio to leverage open
source machine learning libraries to create machine learning-based applications.
The IBM Cloud data platform is designed to help customers break down data
Open Source and most
silos across their own departments and across all of the various data sources they importantly Linux and
need to manage their business. The fact that the platform is based on containers, Kubernetes play a pivotal
microservices, and APIs enhances the ability to make data services become an role in IBM’s transformed
equal citizen in creating a scalable hybrid and multicloud environment that is cloud strategy. In light of
imperative to ensuring that the hybrid and multicloud environment can support the concentration on open
the goals of the business. source, it is not surprising
that IBM would acquire Red
Hat as a natural partner to
Leading with Open Standards to Support the Scalable Hybrid Cloud
execute on its strategy.
IBM has been on a journey to create a hybrid, multicloud environment to
support the need of its enterprise customers. Open Source and most importantly
Linux and Kubernetes play a pivotal role in IBM’s transformed cloud strategy. In
light of the concentration on open source, it is not surprising that IBM would
acquire Red Hat as a natural partner to execute on its strategy. The combination
of an adherence to standards and open source and the focus on enterprise
scalability should position IBM well for the future.
Hurwitz & Associates is a strategy consulting, market research and analyst firm that focuses on how
technology solutions solve real world customer problems. Hurwitz research concentrates on disruptive
technologies, such as Cloud and Multicloud Management, Big Data and Analytics, Machine Learning and
Artificial Intelligence, Information Security, DevOps, and emerging infrastructure platforms that are defining
the future of business agility. Their experienced team merges deep technical and business expertise to
deliver the actionable, strategic advice clients demand. Additional information on Hurwitz & Associates
can be found at www.hurwitz.com.
All rights reserved. No part of this publication may be reproduced or stored in a retrieval system or
transmitted in any form or by any means, without the prior written permission of the copyright holder.
Hurwitz & Associates is the sole copyright owner of this publication. All trademarks herein are the property of
their respective owners.
www.hurwitz.com