Towards a Novel Metadata Information Service for Distributed Data Management

The trend in Grid computing towards more data intensive applications, accessing more and more relational databases and requiring advanced integration of secondhand and publicly available data sources, is still upstanding. Rich metadata information about these data sources plays a vital role for efficient distributed data management. There is a lack of service oriented monitoring tools providing a rich set of metadata for data sources, in contrast to the established usage of monitoring and metadata information tools for network and computational Grid resources. We allow the improved integration and exploitation of data sources in service-oriented architectures by providing a service oriented monitoring tool, named DSMON, for them – initially focusing on relational databases. Our approach supports coarse and fine grained information about heterogeneous relational databases via a uniform interface and provides a homogeneous view on the available metadata. This paper presents novel usage scenarios needing a monitoring service as well as important requirements to be fulfilled by such a service. Our research prototype is not bound to a specific data access middleware nor necessarily tightly coupled with other components. Therefore it can play a vital role in any service oriented infrastructure targeted towards advanced data access, integration and management. The functionality and performance of the DSMON prototype is demonstrated for commonly used relational databases such as MySQL, PostGRE and Oracle. DSMON is an essential component of a more general metadata information service, called DSMIS, for data sources providing advanced funtionalities like prediction of the future behavior of a data source and external metadata updates. We introduce its overall architecture and discuss its novel functionalities.