This post is in response to a question about whether data integration by design is database centralization and about business agility.
Click "Read More" to see the rest of this post!
Let me explain the data integration by design methods by analogy. If we look at the technical architectures of the 1970’s, computers were characterized as “stand alone” computers. Most often “floppy diskettes” were passed from user to user as a means of sharing data between computers. Before that, we used magnetic tapes, paper tapes and punch cards all of which are now obsolete because of computer networks. The computer networks that became popular in the late 1980’s allowed any computer to share its resources with other computers that are also on the network. One would not consider the technical architecture of the 1970’s agile.
Today’s disparate databases are much like the “stand alone” computers of the past. Disparate databases are not designed to share their data with other disparate databases. The Extract, Transformation, and Load (ETL) approach is like the passing of floppy diskettes. The disparate data architecture, which is based upon disparate databases, is not agile and it never will be!
The data integration by design methods are used to develop “integrated” databases or to convert disparate databases into integrated databases. Integrated databases provide designed and reliable data access paths between integrated databases. Much the way that computer networks support sharing computer resources, integrated databases share their data with other integrated databases without the need for ETL software. The integrated data architecture, which is based upon integrated databases, is far more agile, efficient, and effective than disparate data architectures.
Data integration by design is not centralization of databases. If we begin with 10 disparate databases, we finish with 10 integrated databases. Integrated databases, while distributed, function in harmony like a single “virtual” database. For more information about the integrated data architecture visit: http://www.strins.com/integrated-data-architecture.html