An intelligent remote file server (network, unix, marsupial)
Kim M. Korner · 1986
Limitations of current disk block caching strategies are discussed. A new model for providing remote file service using knowledge based caching algorithms is proposed. The knowledge based algorithms generate expectations of user process behavior which are used to provide hints to the file server. Surplus resources of the remote file server permit incorporation of these hints into caching algorithms. The research involves gathering trace data from a modified Unix kernel and trace driven simulation of remote file server models. Comparisons are made between conventional, knowledge based and optimal models. Further applications of knowledge based strategies in operating systems are discussed.