A distributed, multi-staged, high-throughput middleware for relational databases

In this paper we present a distributed middleware for relational databases. Its development was motivated by the need of improvements on a set of legacy systems from the automotive logistics and tracking industry. The solution proved to be effective once it was possible to increase the throughput of handled Requests and offer a minimum level of fault tolerance by employing a pipelined architecture along with distributed data structures. Its deployment ensured that the same legacy applications ecosystem could evolve and operate under growing commercial demand.