eBook

Preparing for a Data Center or Cloud Migration

To better meet business goals, companies are accelerating their plans to transform their
IT. IT is expected to move thousands of workloads from one or more data centers or
clouds, while maintaining downtime requirements for each application. They deploy
automated tools to move mobile workloads at scale, but they fall short with large,
complex applications. Lacking any awareness of application interdependencies, these
tools leave IT in the dark and at risk for causing a service outage.

The key challenge lies in keeping not only IT, but all project stakeholders across the
enterprise, informed and working together smoothly toward the same goal - despite their
disparate roles and project responsibilities.

Of course, you need a solid methodology that covers discovery, analysis, planning and
execution. But this alone is not enough to assure success. If you are moving just a few
virtual machines over the weekend, it’s really no big deal. However, when moving
hundreds or even thousands of applications and the associated infrastructure, then
managing the complexity and interdependencies while minimizing the business impact
becomes a daunting endeavor.

T D S | P R E P A R IN G FOR A D A T A C E NT E R OR C LO U D MIG R A T IO N 1
508.625.3030 www.tdsi.com
 On the surface, most migration methodologies appear to be
similar. The core phases of initiation, discovery, analysis,
Before kicking planning, execution, and maintenance are pretty standard.

off a migration The devil is in the details, however. Or more appropriately, the
project, it is devil is in how the details are managed and orchestrated.
In large-scale data center transitions and migration projects, it’s
crucial to have a critical to have a firm grasp of the environment’s
complete and interdependencies and an understanding of the impact when
holistic view of those dependencies are broken.

the project. How can you plan and manage these phases with minimum
impact to application users? Perhaps more importantly, how do
you avoid burning out your subject matter experts who already
occupy full-time jobs and responsibilities?

Before kicking off a migration project, it is crucial to have a complete and holistic view of the
project, while anticipating each step along the way. At the outset, stakeholders must agree
on the assumptions related to resources, capacity, bandwidth and timelines. Any unrealistic
expectations on the size and scope of the project must be clarified with the business.

Deciding on a focused, universally agreed program management approach is important

from the very beginning. No single approach is right or wrong, but the tools you use for
program management should be understood and accepted by everyone on the team.

Enhanced Discovery
It is generally recognized that establishing a holistic understanding of your environment of
applications, assets and interdependencies is an essential first step in developing an
effective IT transition plan.
Regardless whether this is a physical, virtual, cloud or hybrid migration – or planning and
analysis for resiliency, compliance and risk management -- all projects basically begin with
an inventory discovery and validation phase.
A successful discovery phase consolidates information from multiple available sources of
information. As with any data aggregation initiative, the information must be normalized,
validated and transformed into an actionable “vault of truth” to allow for predictable change
management.

T D S | P R E P A R IN G FOR A D A T A C E NT E R OR C LO U D MIG R A T IO N 2
508.625.3030 www.tdsi.com
Most organizations store data across business silos in a variety of tools, databases, and
files but no one person or team has the complete picture of an entire, hybrid IT
organization. Consolidating data into a single source should make it consistently across the
enterprise. The information to be gathered is generally spread across different systems and
sources. Much information can be pulled from existing systems of record. This can include
more global systems (CMDBs and ITSM are some examples) as well as silo-specific
management tools.

Another source of information can be pulled from auto-discovery systems (Vsphere, iQuate,
RISC, etc.) which can be an excellent source of information to go into the “vault of truth” as
part of the overall discovery process.

The third critical source of information is human intelligence pulled from “subject matter
experts”. Human intelligence is especially important for information that is not typically
available from existing systems of record not accessible by auto-discovery.

T D S | P R E P A R IN G FOR A D A T A C E NT E R OR C LO U D MIG R A T IO N 3
508.625.3030 www.tdsi.com
 This may include data related to business
requirements and technical “tribal knowledge”
related to resiliency, security, performance and
compliance and risk management.
QUESTIONS TO CONSIDER
This “enhancement” of the discovery process is
• HAVE YOU CHOSEN A one of the cornerstones of a proven methodology
CENTRALIZED PLANNING &
that we have been using over our 17 years of data
center and cloud migration projects.
COLLABORATION TOOL TO
MAINTAIN A VAULT OF TRUTH
Just like any “big data” type of exercise,
FOR THE ENTIRETY OF THE
information that is being gathered from various
PROJECT? sources will need to be filtered, normalized and
validated so that it can be relied upon to be
• SHOULD YOU MANUALLY consistent and actionable.
CAPTURE APPLICATION
INVENTORY AND THE As you build your information base, visualization
SUPPORTING ARCHITECTURE tools will be very handy to quickly see where gaps
OR SPEND THE TIME AND exist and understand how to move forward.
MONEY TO RUN AN AUTO
DISCOVERY TOOL? By examining what data points are known and
validated, and understanding how the
• ARE YOU CAPTURING ONLY dependencies interact, you will be able to clarify
THE DATA POINTS REQUIRED what data points are either missing or require
TO PLAN AND EXECUTE THE further investigation.
MIGRATION OR ARE YOU
If there is a server that exists in the environment
GETTING BOGGED DOWN
that cannot be tied to anything, this will be a red
COLLECTING UNNECESSARY
flag for further investigation and resolution prior to
DETAILS?
the move.

Have prior applications already been retired and

the server remains operational despite serving no
workload? Or do you just have an incomplete picture and require more investigation?

The value of visualization is to provide clear guidance as to what else is needed to create
the full picture, and to create a comprehensive move plan. Imagine how easy it would be to
miss key relationships and dependencies trying to parse a spreadsheet or database table.

T D S | P R E P A R IN G FOR A D A T A C E NT E R OR C LO U D MIG R A T IO N 4
508.625.3030 www.tdsi.com
Analysis and Planning
Once your environment has been discovered, documented and validated, the analysis and
planning phase requires strategic decision making based on how the data center migration
will impact your organization throughout the project.

• Is this going to be a “big bang” event where you move everything all at once, say
over a weekend?

• Or, will there be multiple move events where assets are divided into bundles, and
therefore understanding all the interdependencies of an application is critical?

Typically, various groups in an organization will have different outage tolerances and the
plan must take these variances into account.

For example, there is less tolerance with system outages affecting end customer
satisfaction compared to various internal-use applications that are not as mission critical.

In the planning stage, it is important to analyze how your specific business or industry
requirements may affect decisions in terms of project timeline, outage windows and
business impact. The use of planning tools that are iterative and progressive in nature can
be invaluable when future steps like minute-by-minute runbooks come into play.

QUESTIONS TO CONSIDER

• DOES THE TEAM HAVE A DEEP UNDERSTANDING OF THE ENVIRONMENT TO BE

MIGRATED?
• IS THERE A VISUALIZATION OF APPLICATION, NETWORK AND SYSTEM
DEPENDENCIES?
• DOES THE TEAM FULLY UNDERSTAND THE PROJECT OBJECTIVES?
• HAVE ALL THE COMPETING PRIORITIES BEEN BALANCED, INCLUDING SCHEDULE,
BUDGET, MAINTENANCE / DOWNTIME WINDOWS, BUSINESS IMPACT AND MOVE TEAM
CAPACITY?
• HAVE THE RUNBOOK AND PRIORITIZED MOVE TASKS BEEN MAPPED TO A TIMELINE
TO VALIDATE THE MOVE WINDOW IS ACHIEVABLE?

T D S | P R E P A R IN G FOR A D A T A C E NT E R OR C LO U D MIG R A T IO N 5
508.625.3030 www.tdsi.com
Execution
Timely communication, fresh data, clear task
ownership and full visibility throughout the
execution process will significantly improve your
chances of success. That’s obvious.

But how do you achieve this?

Traditional methods for communication and

collaboration based in spreadsheets, Microsoft
Access and Microsoft Project are generally too
labor intensive, single threaded and fragile to
work efficiently in a team environment where
success is measured in hours and minutes, not
days.

Your complete migration can now be

choreographed with fully documented and day-
by-day and minute-by-minute runbooks. These
can also be built so they adapt to new
information and new requirements which
inevitably come up along the way.

Imagine its two weeks before the migration

event and new business or technical information suddenly arises and will likely impact your
plan; how do you respond? It’s very common to receive notice of an addition or deletion in
the week prior to a move event. Can you easily factor in the change and continue with your
plan without missing a heartbeat?

At times like this, automated runbooks can empower the team to be responsive without
introducing risk or working through the night validating all the revised plan elements.

Fast forward to Move Day and picture a typically loud, chaotic “war room” during the event
where people on a conference bridge are talking in a serial fashion, following commands,
and assuming everything is well coordinated from the top.

T D S | P R E P A R IN G FOR A D A T A C E NT E R OR C LO U D MIG R A T IO N 6
508.625.3030 www.tdsi.com
 Most human error occurs when task owners
either do not have the most up-to-date
information from the project managers or have
missed an important detail from a key planning
meeting.

For example, when a team member is just

following orders on an outdated spreadsheet—
and shuts down a server at the wrong time—that
person’s error can delay the project by several
hours if a database now requires a rebuild or
restoration.

To delve even deeper, think about the traditional,

manual methods of command and control during
a migration event such as email, phone calls, and
spreadsheets. These obsolete approaches bring
about unnecessary interruptions and delays that
will drastically slow down progress.

Status updates can also be disruptive as stakeholders constantly check in on status,

consuming valuable resources while trying to alleviate their own concerns that all steps are
progressing smoothly.

Consider instead, using a collaborative, real-time website or dashboard with information

visible to all appropriate levels of management and IT.

Having access to up-to-the-moment information minimizes disruptions to the team, keeps

everyone moving forward and o -track, while allowing stakeholders to remain fully informed
of the progress at their pace and discretion.

Migration events are commonly described as controlled disasters, but maybe it’s time to
raise the bar and expect excellence. Anything can happen during a data center migration
event—from a weather event or power outage to a network or storage failure.

Not realizing or planning for these common risks ahead of time can be the downfall for
those inexperienced in migrations. Before a move event occurs, simulating the step-by-step
process with every member of the move day team is invaluable to the success of a move.
This will ensure readiness of required resources and potentially identify planning gaps
related to any late changes. ideally this simulation should be as close as possible to the

T D S | P R E P A R IN G FOR A D A T A C E NT E R OR C LO U D MIG R A T IO N 7
508.625.3030 www.tdsi.com
actual “move day” circumstances. the old phrase certainly applies here, “You’re only as
strong as your weakest link.”

During the tabletop exercises, IT can double check that task assignments, event durations,
and other move factors. Depending on the systems used, you may even catch errors and
correct them on the spot by using runbook automation if offered by your selected toolset.
This rehearsal ensures the runbook is correct on the move day because IT has validated
and practiced everything ahead of time.

A migration is much like moving into a new home. Everything is clean, organized, and
landscaped just the way you want it on day
one.

However, it can fall into disarray and disrepair

if you fail to do regular maintenance. In
preparing to move to your new data center
“home” or cloud destination, you have
cleaned and validated the data. It is well
mapped out and documented.

Now you should consider implementing a

robust maintenance plan to ensure your
environment will continue to run smoothly
over time, as well as be ready for future
updates and upgrades.

The tools and processes you have used for all

phases of the data center migration can
provide ongoing value on a continuous basis
to support maintenance best practices. Don’t
let all the discovery and cleanup from the last
several months go to waste.

TDS is here to help.

With close to two decades of experience managing complex IT migrations, combined with
the orchestration power of TransitionManager, TDS is empowering the wave of mass
hybrid cloud.

T D S | P R E P A R IN G FOR A D A T A C E NT E R OR C LO U D MIG R A T IO N 8
508.625.3030 www.tdsi.com
 Let us show you how we can help your organization manage
Learn More your data center or cloud migration.

T D S | P R E P A R IN G FOR A D A T A C E NT E R OR C LO U D MIG R A T IO N 9
508.625.3030 www.tdsi.com