What is Distributed Programming?

For this article, distributed programming means using different autonomous computers connected over a network to solve a single problem. Such a hardware configuration is called a multi-computer; these computers don't have direct access to each other's memory and peripherals. Programming multi-computers requires models which are different from normal systems and in which the programmer can transfer data between different parts of the program through shared memory, either in global variables or in the stack.

Among the reasons for writing distributed programs are the following:

• To achieve more processing speed by using more than one computer at a time.

• Having more than one computer could improve reliability. If one computer fails, others might be able to compensate for the loss.

• Some problems, like remote data acquisition, require a distributed environment in which to work.

• It should be possible to start processes on remote computers.

• The necessary data for these processes must be provided before they can do any work.

• Some mechanism for synchronizing these processes should be available, so that they know when to access the data, and produce any results.

Linux already has some mechanisms for processes in the same computer to exchange data and synchronize between themselves. This is called Inter-Process Communication or IPC. One prominent example is the System V IPC, first introduced in AT&T's System V UNIX. This is a set of mechanisms, which includes:

• Shared memories can be shared by more than one process and can be read or written at any time. When one process writes some information in a shared memory, other processes can immediately use it.

• Messages are data structures which can be exchanged between different processes. Unlike shared memories, sending and receiving messages requires the use of system calls.

• Semaphores can be used by different processes for synchronization purposes.