Complex computer systems often imply complex coordination of distributed components that are prone to failures and reconfiguration in the course of their regular operation. This paper describes a communication problem, when communication takes more time than allowed. One solution to the problem is given by introducing an additional middleware layer. In the case of communication timeout, this middleware allows the application to continue its operation and promptly react to this type of error. Middleware design is presented and in a test example functionality of this approach is tested. The analysis of the results shows the limits of this approach and the directions for further research.
