Performance evaluation of UML software architectures with multiclass Queueing Network models

Abstract

Software performance based on performance models can be
\napplied at early phases of the software development cycle to
\ncharacterize the quantitative behavior of software systems.
\nWe propose an approach based on queueing networks models for performance prediction of software systems at the
\nsoftware architecture level, specified by UML. Starting from
\nannotated UML Use Case, Activity and Deployment diagrams we derive a performance models based on multichain
\nand multiclass Queueing Networks (QN). The UML model
\nis annotated according to the UML Profile for Schedulability, Performance and Time Specification. The proposed
\nalgorithm translates the annotated UML specification into
\nQN performance models, which can then be analyzed using
\nstandard solution techniques. Performance results are reported back at the software architecture level in the UML
\ndiagrams. As our approach can be fully automated and uses
\nstandard UML annotations, it can be integrated with other
\nperformance modeling approaches. Specifically, we discuss
\nhow this QN-based approach can be integrated with an existing simulation-based performance modeling tool.