next up previous contents
Next: Resource sharing Up: Introduction Previous: Characterization of distributed systems   Contents

Examples (historical overview)

One of the simple distributed systems are systems in our network:

Figure 2: simple distributed system
\begin{figure}
\begin{center}
\epsfig {file=ds02.eps,height=10cm} \end{center}\end{figure}

Before having distributed systems there existed multiuser systems built around a single time shared computer. UNIX is one of the best-known operating systems for multiuser systems.

Figure 3: centalized system
\begin{figure}
\begin{center}
\epsfig {file=ds03.eps,width=12cm}\end{center}\end{figure}

Since UNIX was already in widespread use and easily available, a lot of distributed system researchers extended UNIX as a basis for distributed systems. The first extension, developed at the University of California at Berkeley (BSD) in the late 1970s, was interprocess communication.
The distributed UNIX model outlined in figure 2 above is the result of the extension contributed by Sun Microsystems: the remote procedure call (RPC) that led to the development of the network file system (NFS) and network information system (NIS). Although the development of distributed UNIX systems offers obvious commercial and technical advantages, UNIX was originally designed to meet a restricted range of requirements in a time when the limitations of computer hardware restricted the scope of the designers' ambitions. In the recent past a new generation of distributed operating systems has been developed. Mach, Chorus and Amoeba are operating systems which have been developed as true distributed systems.


next up previous contents
Next: Resource sharing Up: Introduction Previous: Characterization of distributed systems   Contents
Prof. H.P. Oser
2001-06-08