|
Add Structure to Unstructured Legacy Data
A 10 minute step-by-step white paper
by NetManage, Inc. and HiT Software, Inc.
Introduction
Industry figures quote that at least 60% of
corporate business data and logic is still stored in “legacy”
systems, such as mainframes, AS/400s, and UNIX systems. An even
more impressive statistic is that all the COBOL (running on Mainframe)
daily transactions exceed the total daily number of hits on the
Internet. These COBOL transactions represent the vast majority of
business transactions and therefore, business data.
Multiple methods of tapping into the data
behind these transactions exist. Some complicated, some not so
complicated. Methods from non-intrusive screen scraping to direct
WebSphere MQ integration, to Web services being called directly from
CICS.
The bottom line is that when it comes to
exposing data locked within legacy systems, the data is actually
being moved out of one system, and into another to preserve or
create structure where there once was none.
Most organizations tend to overcomplicate
this integration. Some organizations try to unload and migrate
all the information from legacy applications to a more modern
packaged application like SAP. What some organizations fail to
realize is that the legacy application has been molded and has
evolved over the past 10 years to fit their exact needs. The CIO
expects SAP to fit into the organization overnight which is
almost always an impossibility. Additionally, some organizations
expect the performance to remain on par with the legacy system.
There is a reason that these legacy systems have survived for
so long: they are built to perform.
Another subset of organizations decide to
pump and dump the data to work in conjunction with a new packaged
application. This involves either a massive export/import process
nightly or involves bus-based integration like with JMS or
WebSphereMQ. If the IT organization’s pockets are deep enough and
there is enough collective will within the organization to force a
pump and dump solution, it will eventually work out.
In reality, when it comes to using legacy data
and transactions for use outside the legacy system, the most
realistic method used by organizations is the structured real
time export of typically unstructured data: the organization decides
to keep the legacy system as the primary system of record while
enabling external data users and business units with structured
data via relational database (RDBMS). This is a win-win for
the organization since external users like analysts or accountants
have access to the data in a meaningful way and the primary users
of the system continue to run business as usual.
The fastest, easiest, and most advanced
way of exporting this data in real time with structure is to use
two technologies from two industry leaders.
The first part of the overall solution
is from NetManage and it is called OnWeb. OnWeb is a full featured
legacy integration server that is capable of moving green screens
to the Web, combining legacy transactions into Web services, and
encapsulating and self-describing unstructured legacy transactions.
OnWeb integrates legacy transactions for some of the largest
companies on the Fortune 500.
The second part of the overall
solution requires specialized technology in both mapping
and databases. When it comes to providing a solution that
requires strong skills in these specialized areas,together
with a large market presence and experience, there is none
better than HiT Software, Inc and their Allora database mapping
product. Allora leverages leading edge XML mapping and database
technology to give application developers bi-directional access
to relational databases without the need for complex SQL or
XSLT programming.
Combining these two products, OnWeb and
Allora, an organization can, within 10 minutes and without writing
any SQL statements, export unstructured legacy screen data into
any structured RDBMS in real time.
This whitepaper demonstrates that capability
and outlines the process.
Next: Conclusion
Back: White
Papers
|
