The high level design of the Distributed Computer System created and implemented at the University of California, Irvine is described. This is the gross design for a reliable, fail soft system utilizing low cost processors. The desire to combine reliability and low cost leads to the design of a computing system with a network architecture. The important features of this system are the distribution of hardware, software, and control among the components of the network. This distribution is enhanced by the use of a ″communication by name″ communications protocol rather than addressing the processors of the system directly because the system is both redundant and modular, it is relatively immune to total failure and, in those instances where error or failure does occur, exhibits fail-soft behavior.
|ジャーナル||Bull Sci AIM|
|出版ステータス||Published - 1975 1月 1|
ASJC Scopus subject areas