What Is Distributed Database Systems?

           In many large computer systems, DBMSs are distributed. A distributed database simply divides data among several small computers connected through a network instead of consolidating them in a single database on a large centralized mainframe still the most widespread practice for storing massive databases in large companies. For instance, customer-addresses information may be stored at a cooperate office site, while credit histories may be stored in the credit department, which is located in another building across town. In a distributed DBMS, data are divided so as to optimize performance measures such as communications cost, response time, storage cost, and security. Moreover data can be placed at the sites at which they are most needed and best managed.

           A user calling into the database generally will have no idea where the data are coming from; they could be stored in a computer system in the same building, in a different state, or even in a different country. When the user makes a request to a distributed DBMS, it is up to the DBMS to determine how to best get the data; how it will do this is transparent to the user. Ideally a common seamless interface exists among member systems. Thus, users should be able to work under the illusion that all the data are stored locally. Consequently, the user at any client workstation need learn only one set of rules to get information from any server hooked into the network, resulting in minimal delays.