Adaptive and distributed query evaluation framework for query evaluation in multi-scale contexts

Victor Cuevas-Vicenttin
under the direction of Genoveva Vargas-Solar and Christine Collet
funding: ANRWebContent (Jan 07 - june 09)

Distributed query execution is a problem that has been widely studied. This subject remains important and still introduces challenges whenever it is necessary to process large quantities of data which are highly distributed. Efficient query evaluation implies the utilization of appropriate distributed query evaluation techniques. A query evaluation system should function over a large number of data sources and must manage queries dealing with large volumes of data. This process must take place without affecting the evaluation performance in general and without demanding excessive computation capabilities from the participating servers. The evaluation of distributed queries must therefore adapt to the architecture in which it occurs, for example p2p, grids, etc. For distributed query evaluation we must also use techniques appropriate for the characteristics of the involved resources: capability and disponibility of the data sources, and computation capability of the servers.

The objective of this thesis is to propose a framework for distributed query evaluation that allows developing cooperative query evaluators capable of dynamically adapting themselves to their execution context.