Extending SMART2 to Predict the Behaviour of PL/SQL Applications

Information Systems are nowadays becoming more and more complex and the use of the parallel and distributed approaches has introduced new difficulties in trying to guarantee the performance of end-user applications. In addition, the increasing volume of data to be stored, managed and processed requires to have a well-tuned database and well-designed-and-optimised database programs. SMART2 was developed as an initial solution to cop with the problem of predicting the performance of SQL-based relational programs. In this paper an extension of the SMART2 workbench for predicting the performance of PL/SQL applications is presented. The proposed tool provides end-users with a graph-oriented formalism for specifying the essential characteristics of the application he/she is conceiving and capabilities for matching software components and operational environment.