Distributed System
Distributing the computation among the several processor is called a distributed system. Multiple processor does not share the memory or clock because they have their own local memory. Communication among the multiple processors is done by using the various communication lines such as high speed buses or telephone lines. These kind of systems referred as loosely coupled or distributed system. Processor may vary in size and function in distributive system. They may include workstations, small microprocessors, general purpose computer and minicomputers. The processors are referred by different names such as nodes, sites, computers and so on depending on the context in which they are declared. There are various reasons to use the distributed system:
- Resource sharing
- Reliability
- Computation speedup
- Communication
Resource sharing
In case if there are a number of different nodes are interconnected then user at one node may be able access the resources that are available at another node. Let’s take an example a user at node X may be using the printer which is available at node Y meanwhile the node Y is using a file which locates at node X. Generally, resources sharing in a distributive system offers mechanisms for processing information in distributive database, sharing files at remote nodes, using remote specialized hardware devices, printing remote files, and processing other operations.
Reliability
In case if a node fails in a distributive system the remaining nodes can potentially continue operating. If the system is composed of a number of large autonomous installations, then the failure of one of them will not affect the rest. If satisfactory redundancy exists in system, both in data and hardware then the system can continue with its operations even some of its nodes have failed
Computation Speed Up
If a particular computation may be fragmented in to a number of subcomponents which can run synchronously then this system may allow to the computations among several nodes to run that computation concurrently. In case of specific node is currently overloaded with jobs, some of them can be moved to other lightly loaded node.
Communication
There are many cases in which programs needs to exchange data with one another system. For example, windows system, since they often transfer or share data between displays. In communication network when there are many nodes connected by one another the processes at different nodes the opportunity to exchange information. Users can communicate or initiate file transfer with one another via electronic mail.