Continuous Performance Testing in Virtual Time

In this paper we show how program code and performance models can be made to cooperate seamlessly to support continuous software performance testing throughout the development lifecycle. We achieve this by extending our existing VEX tool for executing programs in virtual time so that events that occur during normal execution and those that occur during the simulation of a performance model can be scheduled on a single global virtual time line. The execution time of an incomplete component of an application is thus estimated by a performance model, whilst that of existing code is measured by instrumentation that is added dynamically at program load time. A key challenge is to be able to map some or all of the resources in a performance model to the real resources of the host platform on which the application is running. We outline a continuous performance engineering methodology that exploits our unified framework and illustrate the principles involved byway of a simple Java application development case study.