ABSTRACT
This chapter presents the guidelines and high level tasks required to finalize the distribution strategy. It looks at the design of the replication strategy, including why replication is used and the various forms of replication available. Implementation of a vertically segmented data store may lead to a more complex strategy for maintaining referential integrity because the data for a single record is distributed around the corporation. Distributing the processing almost always means that the processes must communicate, which implies some form of middleware or distribution services, which, in turn, implies complexity. At the initiating or sending site, each application is designed to perform two main functions. First, the application must perform all the processing required to update the database with the new data. Second, the application must decide if the updated data should be distributed to other remote sites, then distribute the updated data or trigger their distribution.
