Some Thoughts on Data Sharing Among Autonomous Cooperating Database Systems
As our society becomes increasingly information dependent, there will be ever greater pressures for information sharing among separate entities. Organizations will find that, in order to cooperate effciently, they will have to share data with their partners. The rapid spread of networking technology will speed this process along, since it will become more convenient, faster, and less expensive than ever to communicate with others. However, although the benefits of cooperation are self-evident, it seems clear also that many organizations will not be willing to surrender autonomy over their data as the price of cooperation. In this paper we describe our current thoughts about the issues involved in sharing information effectively among a very large number of independent database systems. A major point of our work is to preserve the autonomy of the participants while still allowing efficient sharing. This work has as a starting point the concepts behind federated databases, but we are expanding these ideas to apply to systems with much larger numbers of participants than usually considered, supporting very high transaction rates,
and allowing for a greater degree of heterogeneity than in previous research.