Google Cloud Whitepaper

May 2020

CIO's Guide to
Application
Migration
Table of Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 01

Chapter 1: Where to begin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04

Building an application inventory

Chapter 2: Should I migrate my application to the cloud? . . . . . . . . 07

Chapter 3: Which migration path is right for me?. . . . . . . . . . . . . . . . . . . . . 09

Migrating to GCP managed services

Migrating to containers on Google Kubernetes Engine (GKE)

Migrating to VMs on Google Compute Engine (GCE)

Migrate to Google Cloud VMware Engine

Chapter 4: Modernization without Migration. . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Anthos on-premises

Chapter 5: Project implementaton. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
 1

Introduction
A well-designed
Increasingly, enterprise organizations seek to leverage the many digital
advantages delivered by a successful digital transformation. At transformation
Google Cloud, we believe that effective integration of public cloud
capabilities is fundamental to this critical journey. By augmenting or
strategy should do
replacing on-premises infrastructure with cloud resources, much more than
enterprises can eliminate many of the IT-centric constraints that often
keep you
distract from, and sometimes inhibit, their core business. Cloud
integration opens the door to simplified IT management, elastically competitive...instead,
scalable resources, flexible cost models, and innovative new it should position you
services...ultimately enabling increased operational efficiency and
fueling business growth.
to excel by
untethering IT staff
In addition, as cloud resources have become broadly and easily
from low value,
accessible, they enable many organizations to proactively evolve their
IT infrastructure to support continued success in increasingly labor-intensive tasks,
competitive markets. Sometimes, by maintaining an unoptimized allowing them to
status quo, organizations can inadvertently hamper innovation and
squander internal capital, thereby increasing exposure to market
focus on innovation
disruptors and aggressive competition. and high-impact
projects.
However, a well-executed digital transformation should do much
more than keep you competitive...it should also position you to excel
by untethering IT staff from low value, labor-intensive tasks, allowing
them to focus on innovation and high-impact projects. Also, replacing
(or supplementing) legacy systems with modern technologies can
reduce complexity and cost, while also positioning you to leverage
cloud-native tools to achieve enhanced business intelligence and key
strategic insights. Finally, with nearly unlimited scalability at your
fingertips, applications can scale up and scale down on demand,
while you pay only for what you consume. This allows you to maintain
a continuously right-sized cost profile, while also accelerating
development and reducing procurement cycles. Each of these
benefits can yield tremendous value to your business which, after all,
is the goal of a digital transformation. Unfortunately, while many
organizations are ready to embrace a digital transformation via the
cloud, it can be challenging to devise an effective migration strategy.
They need practical, prescriptive approaches, designed in context
 2

with their technical requirements, business goals, and a deep

knowledge of cloud technology... thereby enabling them to reap the One great way to
cost, efficiency, agility, and innovation benefits of cloud in a way that
achieve this is to
returns value continuously. We’ve written this paper to help. Here, we’ll
provide simple, prescriptive guidance to assist with the most partner with a cloud
important part of your digital transformation: the beginning. provider who will
work with you to
In this paper, we’ll focus on migration of on-premises datacenter
infrastructure. However, similar guidance would also apply to
help:
infrastructure hosted elsewhere (e.g. in public clouds or in an
on-premises colocation facility) and it’s worth noting that migration
1) Assess your
into Google Cloud from other clouds can offer a quick, simple path to
cost savings and enhanced flexibility.
environment

We’ll first explain how to perform effective discovery and assessment 2) Plan your cloud
to ensure a clear understanding of your current application stacks architecture and
and underlying infrastructure. Then we’ll help you leverage that
information to define the first steps for transforming your existing
migration roadmap
applications and workloads. Since different applications may be best
served by differing strategies, those first steps can be crucial to 3) Perform the
success. migration
Truthfully - there is no single right answer when embarking on digital 4) Ensure efficient
transformation and planning a corresponding migration strategy.
Every transformation will have its own nuances and unique
optimization, and
considerations. It’s about understanding the advantages and
disadvantages of the options at your disposal and realizing that you 5) Support your staff
are embarking on a journey. One great way to achieve this is to during and after the
partner with a cloud provider who will help you:
migration to ensure
1. Assess your environment their productivity
2. Plan your cloud architecture and migration roadmap post-migration
3. Establish cloud landing zones

4. Perform the migration

5. Ensure efficient optimization, and

6. Support your staff during and after the migration to ensure

their productivity post-migration.
 3

Digital transformation is truly an exercise in change management, requiring determined effort and focused
leadership across people, process and technology domains; it is as much a culture shift as a technology revamp. But
the rewards in velocity and productivity are certainly worth the journey. Rest assured, we will be with you every step
of the way.
 4

Chapter 1:
Where to begin?

When deciding on a migration strategy, we recommend taking a

comprehensive view, including not only technical requirements, but
also consideration of your business goals (both present and future),
any critical timelines, and your own internal capabilities. Depending
on your situation, any one of these considerations may necessitate a
particular approach. For example, if faced with aggressive timelines
for a datacenter shutdown, you should consider a multi-stage
migration that may include an initial “lift and shift” of application
infrastructure as-is...thus delivering immediate infrastructure
modernization via relocation to cloud. The “lift and shift” can then be
followed by additional modernization at a later time. Alternatively, if
seeking to immediately leverage the full benefits of cloud (e.g.
elasticity, scale, managed services), it may be most efficient to
modernize more aggressively (e.g. by adopting container technology)
in conjunction with migration.
Understanding your
starting point is
In all cases, however, it is important to carefully consider the
essential to planning
characteristics of the applications and workloads that power your
business. It is crucial to carefully assess those applications to deeply and executing a
understand their requirements, their dependencies, and the state of successful
your current environment. Understanding your starting point is
essential to planning and executing a successful application
application migration
migration strategy. strategy.

Building an application inventory

To scope your application migration efforts, you must first catalog the
various items, such as applications and hardware appliances, that
exist in your current environment, along with their dependencies.
Building this inventory is a non-trivial task and we recommend that
you leverage the expertise of the internal teams responsible for the
design, deployment, and operation of each workload in your current
environment, in addition to the teams responsible for the environment
 5

itself. Also, we encourage you to contact us for recommendations Contact us for

and assistance with tools that can help simplify your internal
assessment. recommendations
and assistance with
The inventory should not be limited to applications. Instead, at a
minimum, it should also contain the following:
tools that can help
simplify your internal
• The dependencies associated with each application (e.g.
databases, message brokers, and configuration storage assessment.
systems). In other words, the full application stack
• Any dependencies between applications (e.g. within sequential
workflows)
• Services supporting your application (e.g. source repositories,
continuous integration (CI) tools, and artifact repositories)
• Servers...either virtual or physical and the current operating
systems
• Resource utilization of all compute workloads, VMs,
hypervisors, etc..
• Physical appliances (e.g. network devices, firewalls, and other
dedicated hardware)

When compiling this list, you should also gather salient information
about each item, such as:

• Source code locations and whether you are able to

1) modify the source code and/or
2) rebuild the application
• Deployment methods for the workload in a runtime
environment. (e.g. whether you use an automated deployment
pipeline or a manual one)
• Network restrictions
• Security requirements
• Licensing requirements for any software or hardware
• Critical performance requirements

Your inventory should also include non-technical information, such as

compliance requirements. For example, you should capture the
licensing terms under which you are allowed to use each item. While
 6

some licenses allow you to deploy an application in a cloud

environment, others explicitly forbid cloud deployment. In addition,
some licenses are assigned based on the number of CPUs or sockets
in use, and these concepts might not be applicable when running on
cloud technology. Also, some of your data may have restrictions
regarding the geographical region where it's stored. Finally, some
sensitive workloads can require sole tenancy.

After you complete your inventory, you should analyze the

characteristics of each application stack to help facilitate decision-
making and planning for your transformation journey.
 7

Chapter 2:
Should I migrate my application to Google
Cloud?
Google Cloud offers
The benefits of cloud infrastructure are numerous and we strongly
many options to
recommend leveraging Google Cloud to extract maximum value from
your applications and workloads. Google Cloud offers many options support your
to support your application stacks including Google Kubernetes application stacks
Engine, Google Compute Engine and various managed services
covering storage, networking, big data, monitoring, AI/ML, and more.
including Google
That said, there are situations where migration to cloud may be either Kubernetes Engine,
1) technically infeasible or 2) impractical in the near term. To
Google Compute
determine whether your application can and should migrate to cloud,
begin by asking yourself the following questions: Engine and various
• Are the components of my application stack virtualized or
managed services
virtualizable? covering storage,
• Can my application stack run in a cloud environment while still networking, big data,
supporting any and all licensing, security, privacy, and
compliance requirements? monitoring, AI/ML,
and more.
• Can all application dependencies (e.g. 3rd party languages,
frameworks, libraries, etc.) be supported in the cloud?

If the answer is “No” for any of the above questions, we recommend

evaluating whether it is feasible to replace those application
components with a cloud offering. If not, we recommend leaving
those components on-premises during the initial phase of your digital
transformation, while you focus on the migration of your other
application components. If retention on-premises is no longer viable
(e.g. if you must completely shut down your datacenter) or if you want
to increase proximity to cloud resources, then taking advantage of
Google Cloud’s Bare Metal Solution, or shifting to a colocation facility
(colo) adjacent to the appropriate cloud region are recommended
alternatives. Colocation facilities can allow you to migrate large
portions of the application to the cloud as desired, while still
maintaining high-throughput, low latency connectivity (<1ms) to the
legacy components. Google Cloud also offers a Bare Metal Solution
 8

for running specialized workloads with low latency to other cloud resources. We recommend that non-migrated
components be reevaluated periodically to determine whether migration to Google Cloud has become a more
advantageous strategy.

Application Migration Guidance

Also, note that strategic business considerations must also be factored into any migration decision. As alluded to
previously, if facing a corporate mandate to reduce datacenter footprint by a specific deadline, you may need to
migrate applications more aggressively in the initial phase of transformation. When you engage with Google Cloud to
plan your migration strategy, we will assist you with appropriate scoping, in consideration of both technical and
business factors.
 9

Chapter 3:
Which migration path is right for me?

To help you define an appropriate migration path, we will highlight key

technical considerations and provide prescriptive guidance regarding We suggest a
suggested approaches based on your situation. Of course, as
referenced previously, the migration path you choose should also balanced approach
consider non-technical factors, including, but not limited to: any considering both the
critical time constraints, your available resources, your existing
expertise with cloud technologies, and your appetite for change (e.g.
technical and non-
aggressive vs incremental). We suggest a balanced approach technical
considering both the technical and non-technical characteristics of characteristics of
your business landscape.
your business
As you embark on your transformation journey, we recommend landscape.
considering five key types of migration to Google Cloud:

• Migrating to Google Cloud managed services

• Migrating to containers on Google Kubernetes Engine (GKE) or
Anthos
• Migrating to VMs (“Lift and Shift”) on GCE (Google Compute
Engine)
• Migrating to Google Cloud VMware Engine
• Migrating to the Google Cloud Bare Metal Solution

The guidance below will help you select effective migration paths,
considering both your existing technology stack and the
recommended technology options.
 10

Modernization Guidance for Application Migrations

Migrating to Google Cloud managed services

Commonly-used application stacks are increasingly available in the cloud via managed services. As you begin
devising your migration strategy, check whether managed services are available to support your application. If so,
and if the associated components of your application have minimal customization, leveraging one or more Google
Cloud managed services should be considered as a primary path for migration. In some cases, transition to
managed services may take longer than a wholesale “lift and shift”, however, for many common on-premises
applications, we can provide specific guidance and best practices to assist as you transition to the Google Cloud
equivalent services. Many of these services can be accessed on-demand via Google Cloud Platform Marketplace.

Use of Google Cloud managed services can deliver significant benefits. For example, leveraging these services can
eliminate much of the complexity associated with application maintenance and upgrades, freeing more resources to
focus on your core business. In addition, managed services can also simplify and integrate access to other powerful,
 11

specialized cloud services (e.g. machine learning) that can deliver BigQuery
differentiated value to your business. Within Google Cloud, we have a
wide variety of managed services that may provide value for you,
including: BigQuery, Cloud SQL, Cloud Filestore, and more. Many of
our customers are leveraging these services heavily to benefit their
businesses.

For example, leading global enterprises like 20th Century Fox,

Domino's Pizza, Heathrow Airport, HSBC, The New York Times, and
many others migrated their traditional on-premises databases and
data warehouses to BigQuery for their data analysis needs, helping
them do everything from break down data silos to jump-start their
predictive analytics journey—all while greatly reducing costs.
Google Kubernetes Engine
(GKE)

Migrating to containers on Google Kubernetes

Engine (GKE) or Anthos

If you have application components that are not supported by

managed services, or if you prefer not to use managed services, then
migrating your application into containers on Google Kubernetes
Engine (GKE) should be your next consideration as a first step into the
cloud.

GKE provides a managed environment for deploying, managing, and

scaling your containerized applications and delivers numerous
benefits such as:

• Delivering cost savings by efficiently managing workload

density (i.e. via increased density due to lighter-weight
containers and intelligent workload management)

• Container health monitoring and automatic restarts

• Increasing agility and reducing time-to-market (e.g. by

facilitating faster, automated continuous integration and
continuous delivery processes...a.k.a. CI/CD)

• Access to advanced modern operational management

capabilities with Cloud Operations , Istio and other cloud
services
 12

Some examples of
To determine whether your application is a good fit for
containerization on GKE, consult your application inventory details
good candidates for
and do the following: containerization
include: web
• If your application is developed by a 3rd-party vendor, then
work with the vendor to validate support for the intended application servers,
usage with containers (e.g. dev/test, training, staging, business logic, JEE
production) and to identify any required configuration or setup
updates. Containerized usage supported => Good candidate
middleware (e.g.
for containerization WebLogic,
WebSphere, Apache
• Evaluate the current operating system (OS) licensing model to
confirm the ability to run in containers. If needed, identify any Tomcat), multi-tier
necessary configuration changes, Google Cloud support stacks (e.g. LAMP,
alternatives, and/or GKE-provided OS replacement options.
Compatible licensing => Good candidate for containerization
WordPress), small
and medium-sized
• Assess whether the application workload is resilient to
databases, dev/test
restarts and being scaled out? Resilient to restarts and scale-
out? => Good candidate for containerization applications, training,
labs, and low load
After answering these questions, if your application is a good
candidate for containerization, you should then confirm support for
services.
any prerequisites1 associated with your desired migration. Some
examples of good candidates for containerization include: web
application servers, business logic, J2EE middleware (e.g. Apache
Tomcat), multi-tier stacks (e.g. LAMP, WordPress), dev/test
applications, training, labs, and low load services.

Note that, when leveraged within the Anthos platform, Google-

managed containers can deliver application and workload portability
across hybrid and multi-cloud environments. If a hybrid or multi-cloud
approach aligns with your strategic goals, we recommend adopting
Anthos. With Anthos, you can achieve a unified view of all GKE-
managed environments, thus simplifying global application
management.

1 GKE migration documentation (see “Prerequisites” sections):

 13

Either way, when migrating into GKE, you should consider your long- "Containers are already a
term goals in conjunction with the practical realities associated with
your situation. In the long term, a microservices-based approach will part of our cloud
maximize efficiency in most cases. However, directly adopting landscape, giving us a
microservices may be impractical based on your available resources,
powerful way to
your teams’ expertise, project timelines, etc. As a first step, we
recommend using Migrate for Anthos to lower the complexity of manage and maintain
converting VMs into containers. our systems as well as
As you attain familiarity with containers, cloud, and Kubernetes, we customer environments.
recommend that you eventually consider refactoring your application At the same time, we
stack to adopt a microservices architecture as a next step in your
have a lot of VMs in
transformation journey. Microservices can deliver additional benefits
to your application stack including improved fault tolerance, easier production and we are
maintenance, portability, and enhanced scalability. always looking for
To accelerate your GKE adoption, use the Migrate for Anthos product optimized ways of
to orchestrate automated migration of your applications. As an migrating these over to
example, Atos, a global systems integrator, has been using Migrate
hybrid cloud delivery
for Anthos to accelerate their customer’s cloud journeys.
models. Migrate for
Anthos gives us an
additional fantastic tool
Migrating to VMs on Google Compute Engine
(GCE) in transformational
projects and it will
The vast majority of applications can be migrated as-is to VMs on
further accelerate our
GCE, thus “lift and shift” provides the simplest, fastest path to cloud
adoption. If you are migrating VMware-based workloads to GCE this cloud success.''
would follow a “Lift and Optimize” approach. This would require
changes to your VM operating structure, but would allow you to take
Michael Kollar, SVP for Cloud
full advantage of more cloud-native capabilities. If you want to
Engineering at Atos
modernize in phases, have time constraints, or if your application
stack is not well-suited for immediate containerization...then
performing a “lift and shift” should be your primary path.

Migrating to VMs on GCE provides immediate infrastructure

modernization by delivering access to the scalability and elasticity of
cloud, along with all of the accompanying technical and business
 14

benefits. The broad applicability of the “lift and shift” strategy allows Migrating to VMs on
these benefits to be applied to application components with a wide
range of profiles, including: varying operating systems (e.g. Windows, GCE provides
Linux), demanding performance requirements, specialized kernel immediate
needs (e.g. kernel modules), or with HW-specific license constraints
(e.g. per CPU).
infrastructure
modernization by
A “lift and shift” and “lift and optimize” also set the stage for future delivering access to
application-level modernization. Once your application is in the cloud,
additional modernization (e.g. adoption of container technology) the scalability and
becomes much easier. As you gain familiarity with cloud tools and the elasticity of cloud,
variety of cloud services, you can approach subsequent
modernization efforts incrementally and at your own pace.
along with all of the
accompanying
To accelerate your GCE adoption, consider leveraging the Migrate for technical and
Compute Engine product to orchestrate automated migration of your
applications into GCE VMs. business benefits.

Industry leaders such as Cardinal Health, a $130B Fortune 500

healthcare provider, are reaping major benefits after migrating their
VMs to GCE. Cardinal Health migrated existing systems to cloud to
enhance their security and data protection, to leverage speed and
cost savings, and to gain additional infrastructure flexibility and
business agility. Starting with no prior experience on Google Cloud,
they collaborated with Google Cloud services teams to plan and
execute their migration, achieving production-readiness in just 3
months.

Migrating VMware workloads to Google Cloud

VMware Engine (GCVE)

Finally, if your existing environment is VMware-based, and you prefer

to minimize change to your current operations, GCVE is a managed
VMware-as-a-Service offering purpose-built for VMware workloads
on Google Cloud. GCVE is the fastest, safest, most flexible way to “lift
and shift” VMware virtual machines to the cloud. GCVE enables
enterprise customers to run VMware virtual machines natively in a
 15

dedicated, private Software Defined Data Center within Google Cloud. "Google Cloud has
This allows customers to move VMware from on-premises to Google
Cloud with no customer changes to tools, processes, policies, or been exemplary in
ecosystem integrated solutions. This new service allows our the “how can we
customers to run VMware vSphere-based workloads in Google Cloud,
while also benefiting from Google Cloud's unique strengths, including:
help you” approach
our secure and scalable global infrastructure, our innovative data and with
analytics, and our AI/ML capabilities. Users will retain full, native understanding what
access to the VMware stack including vCenter, vSAN and NSX.
we want to do and
how to make what
we want happen.”

Brian Hanlin, Sr. Director, Cloud

Platform, Cardinal Health

Lift-and-optimize
 16

Chapter 4:
Modernization without Migration

Anthos on-premises

What if you have applications that will have to remain on-premises? In

many cases, those applications may still benefit from infrastructure
modernization. Specifically, leveraging Anthos GKE on-premises can
provide the best of both worlds, enabling the applications to remain
on-premises, while still leveraging the benefits of containerization and
the benefits of Google Cloud. As discussed previously, Anthos
simplifies hybrid and multi-cloud application management, by
delivering a unified view of all GKE-managed environments, whether
running on-premises or in Google Cloud. Finally, by modernizing with
Anthos, you also set the stage for future hybrid cloud integration, if
desired. To learn more about Anthos and to determine whether your
on-premises applications are good candidates for moderization via
Anthos, see this Anthos overview.
 17

Chapter 5:
As you look beyond
Project Implementation
your initial migrations
The guidance above will help define a template for the first steps of to subsequent
your migration strategy, including consideration of non-technical phases of your
factors that may be specific to your business (e.g. risk/timing
associated with migration of business-critical applications). As you
transformation
proceed with both strategic and tactical (including timelines, journey, we can help
resourcing, etc.) planning, we will work closely with you to help you plan accordingly.
achieve your goals. Then, as you look beyond your initial migrations to
subsequent phases of your transformation journey, we can help you
plan accordingly.
 18

Conclusion

Application Migration Guidance Summary

In this paper, we’ve provided simple guidance regarding key decisions required in the first phase of your digital
transformation, for example. whether to migrate applications to the cloud and how to approach those migrations. As
stated previously - there is no single right answer. Digital transformation is a journey and, as with most journeys,
there are many paths to choose from. With the guidance we’ve provided, however, you can begin your journey with
confidence, having defined practical first steps for enhancing your business in partnership with Google Cloud.

If you have questions or would like more detailed guidance based on the specifics of your environment or business
requirements, we encourage you to contact us. We’re here to help.