Year
1977
Abstract
Availability, defined as the ratio of time a computer system is functioning normally to the total time it is in demand, is a measure of the overall accessibility to the system and its operational effectiveness. This ratio allows estimates of down time to be calculated for contingency planning. Single and multiple machine configurations are evaluated in terms of typical component values for mean time between failure (MTBF) and mean time to repair (MTTR), and the Availability of each configuration is computed. Through the use of interconnected, redundant processors with multiprogramming operating systems and all input data provided to each processor, down time for even failure prone units can be reduced from 781 to 32.3 hours per year. Several interconnection methods are shown with projected reliability data.