Date Added: May 2009
Services offered today rely on large amounts of data that can be accessed fast and reliably. One technical solution providing both acceptable speed and high reliability are distributed databases, which can be accessed from an application as one virtual database server. The virtualization here hides complexity introduced by distributing the service to a set of machines. In this paper, the authors will present a DHT-based architecture implementing a lookup layer for such a database, which preserves important features such as self-organization from its DHT roots, but still offers a good performance for time-critical applications. Additionally, first analytical results are given, which show some of the basic mechanisms at work in such systems.