An automated technique for designing optimal performance IMS data bases

ABSTRACT

There exists great potential for IMS performance improvement through selection among design alternatives. One such area is data base design, which covers a number of sensitive variables. The only feasible way to evaluate the effects of varying combinations of values for these variables, is via simulation. Many of the drawbacks of simulation have been alleviated by the creation of a comprehensive, validated model of IMS, which feeds mostly on pre-existing data. This model allows the evaluation of many alternatives, including data base design. A case study has shown the great potential which may be realized through use of this tool.