Adaptive and Reflective Middleware

Middleware platforms and related services form a vital cog in the construction of robust distributed systems. Middleware facilitates the development of large software systems by relieving the burden on the applications developer of writing a number of complex infrastructure services needed by the system; these services include persistence, distribution, transactions, load balancing, clustering, and so on. The demands of future computing environments will require a more flexible system infrastructure that can adapt to dynamic changes in application requirements and environmental conditions. Next-generation systems will require predictable behavior in areas such as throughput, scalability, dependability, and security. This increase in complexity of an already complex software development process will only add to the already high rates of project failure. Middleware platforms have traditionally been designed as monolithic static systems. The vigorous dynamic demands of future environments such as large-scale distribution or ubiquitous and pervasive computing will require extreme scaling into large, small, and mobile environments. In order to meet the challenges presented in such environments, next-generation middleware researchers are developing techniques to enable middleware platforms to obtain information concerning environmental conditions and adapt their behavior to better serve their current deployment. Such capability will be a prerequisite for any next-generation middleware; research to date has exposed a number of promising techniques that give middleware the ability to meet these challenges head on. Adaptive and reflective techniques have been noted as a key emerging paradigm for the development of dynamic next-generation middleware platforms [1, 2]. These