Did you know...
Data integration is unavoidable in an information environment where numerous databases jostle with each other for dominance. Data integration is required to combine different databases and present a unified and seamless view to users.
By Odalys E. Anton
Data integration is one of the most important aspects of information technology today. There are many reasons why it has assumed importance. Firstly, this is required to address the issue of legacy systems. A decade back mainframes ruled the information industry and databases were stored in flat files. As technology matured we moved on to relational database models. Here again , we had small databases like Access working along with giants like Oracle, Sybase and Microsoft SQL server. IBM realizing the potential of desktops, created a version called DB2 UDB to cater to this section of the audience. Web applications added to the complexity of data storage. It is but obvious that a unified platform was required to provide a single window view of all these databases. The cost of converting legacy mainframe databases was too huge and data integration was the only solution.
Another important factor which led to data integration was mergers and acquisitions. Consider a travel portal merging with an airline company. The former may have been working with a MS SQL server and the later with Oracle. Again data integration is the only answer. A unified view of the entire data is required to seamlessly integrate and present information. Accomplishing this is possible with data integration.
One of the data integration techniques widely used is ETL or Extraction, Transformation and Loading. In this method data from various databases is extracted and transformed into a single framework and then loaded in a single database. Users access data from this unified database. This is a simple data warehousing solution and is effective most of the times. The problem arises when a real time database is unified. Since the process of ETL has to be executed on a particular database before making it available to a unified database, users may be provided with stale data unless the ETL process is repeated at frequent intervals. This adds to the overheads and reduces efficiency. Many other models have been designed to address these issues.
The latest development which has given impetus to data integration is SaaS. Many organizations have adopted SaaS tools and this platform cannot perform efficiently without data integration tools. In house databases have to coexist with other vendor databases and seamless operations can be achieved only with data integration.
Home