Design and Implementation of Distributed Databases Using Estelle∗

1 I n t r o d u c t i o n Distributed Systems (DSs) are becoming more and more popular. They have been used to attend the natural distribution of people and information, to provide better performance at lower costs, to facilitate maintenance and to improve the reliability of computer systems. In order to achieve all these objectives DSs have become very complex and today there is no consensus about an exact DS definition. According to Bochmann [1] a DS can be classified having in mind the following types of distribution: ∗ Supported by CNPq, CAPES and FAPESP access, for example, users having access to the services provided by a multi-programming system using terminals at different locations; computers, for example, a large number of users sharing a set of services provided by a computer network; processors, for example, a multiprocessor system where the execution of a task is not done by a single processor but it is distributed among several processors; data, for example, a distributed database system. These types of distribution can be combined, increasing the system complexity, and must be carefully taken into account during the development of a DS. Figure 1 shows the main phases of a Life Cycle of Distributed Systems (LCDS).