As organizations move quickly to adopt Service Oriented Architectures (SOA) as
a way of responding faster to business needs, data must now be treated as a corporate
asset, not just a division or business unit tool. Unfortunately, many data sources
are not in the right format or lack the right metadata to allow quick integration
to be efficiently repurposed for other uses. By implementing an information management
strategy that focuses on data quality and data integration, the benefits of SOA
can be further realized. This approach offers great extensibility, allowing applications
and users to access information not only within the enterprise, but also across
enterprise and industry boundaries. Such complete end-to-end horizontal business
and information integration provides new agility and flexibility to support any
SOA project.
Data quality is becoming an increasingly hot topic as poor data quality undermines
the usefulness of services, lowers user satisfaction and sometimes even breaks
the production systems based on an SOA. To improve data quality, automated data
analysis, including data profiling and monitoring, are indispensable and important
first steps. This hasn't always been the case since many organizations build
data sources to satisfy departmental needs. As applications adapt to constantly
changing business needs, the structure and semantics of the data they create
may change over time. Data format, semantics and rules in one data source are
likely different from other data sources. For example, a finance department
system may store customer data in COBOL format and represent customers from
an accounting point of view, while sales and marketing may store customer data
in a database and define customers from a marketing point of view.
If this organization wants to build a customer master database, the differences
in data format, semantics and rules must be reconciled. This proves to be one
of most challenging tasks in an SOA-based integration. Likewise, mergers and
acquisitions constantly demand additional data integration while data governance
and compliance mandates increasingly create a higher level of integration needs.
Now more than ever, a manual data analysis process is no longer adequate to
solve these overwhelming integration challenges.
Automated Data Analysis and Monitoring
Automated data analysis is important in any SOA project to help streamline data
discovery, reduce the time to analyze and integrate data, and allow users to
understand data sources quickly and determine the suitability for the integration
task at hand. The adoption of data analysis and monitoring solutions as part
of an SOA allows users to automatically scan samples of data to determine their
quality and structure. This, in turn, allows users to build an understanding
of the inputs to an SOA-based integration process ranging from individual fields
to high level data entities. As a result, it helps users more accurately plan
projects and correct problems with the structure or validity of source data
before it impacts projects.
1
Solution Center Resources