If you receive errors when attempting to view this white paper, please install the latest version of
Adobe Reader.
"Informatica's
data migration solution decreases the risk and minimizes the errors inherent in data migration projects.
At the same time, our solution reduces the costs of data migration projects by automating processes,
improving business-IT collaboration, and incorporating proven best practices."
Source : Informatica
Achieving a Successful Data Migration
Data Migration is also known as :
Data Migration,
High Speed Data Migration,
Data Migration Tool,
Data Tape Recovery,
Data Migration Strategies,
Migrate Data Cost-Efficiently,
Data Migration Guide,
Process of Transferring Data,
Data Migration Solutions,
Data Migration Projects,
Data Migration Professional Resource,
Term Data Migration,
Data Migration Pro,
Data Migration Manager,
Data Migration Steps,
Data Migration Process,
Data Migration Testing,
Data Migration Plan,
Migration Information Source,
Transfers Database Schemas,
Data Migration Techniques,
Data Integration Server,
Data Migration Importing Images,
New Data Migration Manager,
Data File Conversion,
Database and Application Migration,
Practical Data Migration,
Database Migration Database Conversion,
Data Migration Resources,
Posts Relating to Data Migration,
Software Data Migration,
Migrate Data Without Disruption,
Easy Migration Tools,
Open Source Data Migration,
Data Migration Options,
Data Integration Tool,
New Best Practices for Data Migration.
Talk to application project owners early in the project lifecycle and you'll hear a lot about the exciting part"the business
process and functionality of the target application. Mention the integration required to migrate data, and they'll probably
downplay its importance. But if you ask the same questions later in the project lifecycle, the emphasis will be
reversed"you'll hear a lot about data migration. The challenges of migrating data become the "tail that wags the dog"
because most data problems are identified too late. Since there is little time left to fix the problem, most application
projects wind up compromised in some way"reduced functionality, budget overrun, or late delivery. For example, if the
application is a new inventory management system that is supposed to reduce inventory by $20 million annually, then
every day that application is late may be costing the business $55,000.
Anecdotal evidence suggests that the data migration phase can consume up to 40
percent of the entire budget of an application implementation, upgrade, or instance
consolidation. And, since there are no separate metrics for migration, data migration
problems can lead an organization to judge the entire application project a failure. The
organization may end up mistakenly concluding that the new software package or the
application upgrade is faulty, when in fact the problem lies in the data migration
process.
Why Migrate Data at All?
Data migration is a necessary part of application migration, the process of moving
business functionality and reliance off one or more existing applications and addressing
the business's requirements with another application. Often the intent is to shut down
the first application (referred to as the source application) sometime after the migration
is complete.
Companies may choose to migrate applications primarily for business reasons, for
example, in order to implement new functionality, consolidate existing and acquired
applications after a merger or acquisition, or outsource to a service provider. They may
also choose to migrate applications to meet the needs of IT, for example, to phase out
applications that are expensive to maintain, or reduce the TCO of their IT environment
by migrating redundant or non-critical applications to a corporate standard.
Most application migrations occur as part of a new application
implementation, such as when an ERP or CRM application is
implemented and existing legacy applications are migrated to the new
system. Two other types of application migrations are
application upgrade, in which the target and source application are
different versions of the same product, and application instance
consolidation, in which the source and target applications are the same
and the goal is to reduce costs by having fewer instances of
the application to run and maintain.
Migration"Strategic or Tactical?
Because data migration is part of a larger, strategic project, most
organizations view it as a tactical challenge, not a strategic project
in its own right. This perspective leads organizations to seek
‘quick-fix' solutions to data migration, but these don't work because
data migration is much more complex than organizations initially
expect, with a longer learning curve than they've usually planned
for. Overlapping tools and technologies on the market compound
confusion about what they need to do. The ‘quick-fix' approach to
data migration ultimately only contributes to the high failure rates of
application migration projects.
Even organizations that have prior experience with migration
frequently fail to leverage their hard-won data migration expertise.
Their ad-hoc approach to migration means that there may be no mechanism
to capture, leverage, or re-use best practices.
In order to improve their data migration success rate, organizations
need to replace a purely tactical approach with a strategic one.
This means addressing data migration, not just as a one-time move, but
as an ongoing process of making the data work"no
matter what changes occur in the company's systems. A data migration
strategy should address the challenges of identifying
source data, interacting with continuously changing targets, meeting
data quality requirements, creating appropriate project
methodologies, and developing general migration expertise.
Use Case Scenario"Move Off The Mainframe
Need: Increase competitiveness by replacing a limited
mainframe-based application with a comprehensive client/server
application.
Approach: Adopt a data integration platform with native mainframe
access to streamline data migration.
Benefit: Increase performance in accessing the mainframe ten-fold,
halve the numbers of resources required to complete the
project, and reduce the number of resources needed for mainframe
maintenance.
Build an In-depth and Up-front Understanding of Data Sources
To create an effective strategy, companies need to dedicate
substantial up-front effort to understanding sources. Simply profiling
and sampling the data are not comprehensive enough to support the
creation of a detailed strategy or an accurate estimate of the
effort required. Instead, organizations need to devote a significant
amount of time to making a full and accurate identification of
source data, including:
- Does it exist?
- Where is it?
- Can disparate data be related?
- What is the focus of each source?
- What about standardization?
- Is sufficient detail available?
- What about unstructured data?
- Is data orphaned anywhere?
Once the source data has been properly identified, the strategy team can then test that the data will support the required
functionality of the target application. The team should begin by identifying quality problems in the source data"such as syntax
and semantic errors, format problems, and integrity issues"and plan how to correct outstanding issues.
The team should also identify and prepare to correct problems
accessing source data. For example, there might be potential
problems accessing legacy data or data from external feeds. The team
may need to address mismatches between the business'
need for timely data and the system's ability to deliver the data on
the business' schedule. Data politics"meaning, issues about
who "owns" certain data"may arise, causing unnecessary delays in
obtaining the appropriate permission to access and cleanse
certain data. And, of course, a good strategy should allow flexible
interfaces to various data sources that can evolve over time.
Move Data to a Continuously Changing Target
Implementing an application is an iterative process, with business
and IT staff making frequent changes to the application
functionality or underlying processes. The tweaking doesn't stop after
implementation, either"throughout the life of an application;
evolving business and technical needs require changes ranging from
minor refinements to major overhauls.
Many of these application changes require adjustments to the data
migration capability. If making those adjustments is complex, it
can increase the TCO of an application to unsustainable levels. The
data migration capability must readily adapt to changes in
business requirements, target interfaces, data models, and data
requirements"before, during, and after implementation.
Ensure Quality
Data quality is critical to the ultimate value of the application.
Bad quality data in the target application will lead to end-user
rejection, turning the target application into shelf ware. But the
problems of poor data quality can cripple an application long before
roll-out. If the data fails to pass a base set of validation rules
defined in the target application, the data load will fail. Bad data in
the
target application can impact business processes after a "go-live" and
necessitate costly manual fixes.
Organizations must establish user confidence in the data. In order
to fully trust data, business users want the ability to trace it, to
find out where it came from and how it was changed. This requires some
form of data lineage capability, such as metadata
management. Data profiling, validation, and cleansing are also
important. In application environments data quality is also about
business rules. Data must obey to unique business rules, as well as
meet validation thresholds.
Use Case Scenario"Consolidate Systems
Need: Simplify parts, materials, and supplies by migrating data from 15 incompatible mainframe and legacy systems to SAP R/3.
Approach: Leverage data integration platform with easy-to-use development tools, native SAP interoperability, and built-in
cleansing capabilities for data migration.
Benefit: Deliver project 25 percent below budget and 85 percent
fewer man years than originally estimated, reduce IT costs
through legacy retirement, improve ability to track costs, calibrate
inventory, and manage storage facilities. Reusable integration
components accelerate future projects, such as additional migrations to
SAP and data warehousing.
Leverage an Iterative, Flexible Project Management Methodology
A data migration process is often doomed from the outset by one
flawed assumption"that migration is a one-time event that
follows a sequential waterfall process of examining source data,
examining the target, designing the transforms, and then building,
testing, and executing the code. Because the source and target are
never what they are expected to be, this sequential process
leads to data problems being discovered late in the project, at which
point they are very costly to fix and cause significant delays to
the overall application project.
Successful data migration calls for a different strategy, one that
enables iterative, flexible development to meet the organization's
changing needs. The best approach"and the only practical one" is driven
by the data and its constraints, with the team designing
as it continues to learn more about the technical issues and business
objectives of the project.
Centralized Resources
When organizations view each application project as a standalone
effort, they often don't realize that some of the expertise that the
team has developed could be applied to another project. This is
particularly true of data migration"teams end up unknowingly
recreating the wheel as they painstakingly build connections to a
source system, unaware that another team has already created
an extensive library of objects for accessing that same source.
Organizations benefit when they centralize their resources to
address all migration projects, encouraging the development and
leverage of migration processes and best practices that incorporate
lessons learned from all projects. A dedicated team finds it
easier to recognize common problems, regardless of business area, and
reuse solutions that they've developed earlier. The team
can also leverage its collective experience of tapping into the same
legacy sources, regardless of different targets. This approach
helps end unnecessary duplication of effort.
But perhaps most important is the shift in strategic perspective"for
a team focused on integration, data migration isn't about
getting one project done as quickly and efficiently as possible. It's
about coming up with the wisest way of managing and leveraging
a key enterprise resource. The team's design decisions, resource
allocations, and prioritizations reflect the needs of the enterprise
as a whole, not just the requirements of the immediate project at hand.
Use Case Scenario"Consolidate and Synchronize
Need: Enhance customer service by putting all customer data from two merged companies in one place.
Approach: Migrate and consolidate 10 legacy systems to Siebel CRM, and synchronize between Siebel and accounting application.
Benefit: Solve more than 97 percent of data quality issues, reuse more than 40 percent of integration processes to accelerate
development, retire legacy systems.
Craft Your Strategy, Choose The Right Tools"And Beat The Odds
Data migration problems often derail valuable application projects
and provoke ‘blame storms' that affect even the most skilled and
conscientious business and IT staff. But it doesn't have to be this
way. Organizations can overcome data migration challenges and
put a stop to this waste by placing a higher priority on data
migration, and acquiring tools or a toolset that supports high-level
system information, technical architecture design, technical process
design, and process rationalization.
Choosing the right tool for this critical task is essential. The
tool should support business acceptance criteria and transition
planning. It also needs to provide key technical capabilities, such as
data re-engineering and cleansing, logical entity mapping, and
logical attribute mapping. Informatica's products support all these
capabilities and best practices for migration, helping hundreds of
leading organizations to migrate data and implement application
projects while meeting timescales, budgets, and data quality
requirements. For more information, please visit www.informatica.com/solutions/technology/migration to view the white paper,
"Accelerate Application Data Migration with Informatica."