Performance MIFs provide a mechanism for automatically moving performance models among modeling tools. Use of the MIF does not require tools to know about the capabilities of other tools, internal data formats or even existence. It requires only that the importing and exporting tools either support the MIF or provide an interface that reads/writes model specifications from/to a file.
Interchange formats have been defined for queueing network models (PMIF+,G-PMIF), software performance models (S-PMIF+), layered queueing networks (LQN), UML, PNs and other types of models. In each interchange format, a file specifies a model and a set of parameters for one run. Since modeling studies typically require multiple runs of the same model using different parameters (e.g. different workload mixes), this requires preparing and exchanging multiple interchange files. In addition, interchange formats do not specify the output metrics that are to be returned after model execution, typically resulting in either a default set of metrics or all possible metrics.
To address these issues, the Experiment Schema Extension (Ex-SE) for defining a set of model runs and the output desired from them was specified. It has the expressive power to specify iterations, alternations, assignments of values, actions based on model results and more. This schema extension provides a means of specifying performance studies that is independent of a given tool paradigm. It requires only that a tool support the Ex-SE or have an interface that is capable of reading/writing extended interchange files.
Experiments generate output metrics, such as response times, utilizations, etc. The experiment specifies the output metrics desired; an Output Schema Extension (Output-SE) specifies the XML format for them. A performance analyst typically studies this output to form conclusions about the results of the experiments, then prepares a presentation and/or report to explain the results. Another extension, the Results-SE specifies how to automatically transform the output metrics into useful results.