Algorithmic Foundation of SkyData

  • Coordinator: S. Devismes (Verimag/UGA).
  • Participants: K. Altisen (Verimag/UGA), L. Arantes (LIP6/SU), T. Begin (LIP/UCBL), A. Beynier (LIP6/SU), E. Caron (LIP/ENS Lyon), S. Devismes (Verimag/UGA), S. Dubois (LIP6/SU), A. Durand (Univ. Clermont Auvergne/LIMOS), E. Jeanneau (LIP/UCBL), C. Johnen (Univ. Bordeaux/LaBRI), F. Petit (LIP6/SU), P. Sens (LIP6/SU), and hired PhD Students 1 and 2.

Description:

The goal of WP2 is to lay out the algorithmic foundations for the SkyData project. This means defining and solving the theoretical problems involved by this new paradigm. This workpackage is the natural continuation of Task 1.3 in WP1, since algorithms will be designed in the models defined in that task. All tasks of WP2 aim at providing algorithmic solutions to the basic services required by Self-Managed Data; those services have been structured in three basic applications. The first one (Task 2.1), called the Replicates Management, handles the global problem of how Self-Managed Data can exist in a SkyData environment, from their creation to their vanishing. This first application is the basis application of the system since no SkyData Environment can run if no data can survive, i.e., appear, live, and perish in the system. The two other applications relate to data usages: Data Exploitation (Task 2.2) and Data Usage Prediction (Task 2.3). Note that Task 2.1 starts at T0+6 and greatly overlaps Task 1.3, so that semantics model can be validated and (if needed) refined or customized according to algorithmic usages. Furthermore, since it deals with the very basis of SkyData, we plan it to last 2 years. Tasks 2.2 and 2.3 will be developed in parallel during the last 18 months of the project.