Investigating different instrumentation techniques in the context of ESB runtime verification

Abstract

With the increasing popularity of the service-oriented architecture, enterprise software is increasingly being organised as a number of services interacting over an Enterprise Service Bus (ESB). The advantages of this arrangement are numerous, not least the modularity and flexibility it affords and the ease with which multiple technologies can be integrated. Yet, the highly dynamic nature of ESBs — where components can be introduced, replaced, or withdrawn transparently — brings with it a number of challenges, particularly to ensure the correct overall behaviour of the ESB. In a flight booking ESB application, such as that depicted in figure 1, back-end airline services may be swapped at runtime, potentially causing unexpected behaviour. The reason for this may be that during the testing phase not all banking services may have been tested thoroughly or that the service experiences an update without the flight booking application being notified.