TANDEM COMPUTERS A Census of Tandem System Availability Between 1985 and 1990

Tandem computer systems are designed to be single-fault tolerant. This paper takes a census of customer system outages reported to Tandem. The census shows a clear improvement in the reliability of hardware and maintenance. It indicates that now (1989) software is the majority source of reported system outages (62%), followed by system operations (15%). This is a dramatic shift from the statistics in 1985. Even discounting systematic under-reporting of operations and environmental outages, the conclusion is clear: hardware faults and hardware maintenance are no longer a major source of outages. As the other components of the system become increasingly reliable, software necessarily becomes the dominant cause of outages. Achieving higher-availability will require improvements in software quality, and software-fault tolerance, and will require simpler operations, and operations-fault tolerance.