Cost Models for Selecting Materialized Views in the Cloud, Application to Amazon EC2 and S3 Services

Data warehouses performance is usually achieved through physical data structures such as indexes or materialized views. In such a context, several cost models enable to select a relevant set of these structures. Nevertheless, this selection becomes more complex in the cloud. Indeed, the criterion to optimize is at least two-dimensional, with the monetary cost balancing the overall query response time. This paper introduces new cost models that fit into the pay-as-you-go paradigm of cloud computing. Based on these cost models, an optimization problem is defined to find, among candidate views, those to be materialized to minimize both the overall cost of using and maintaining the database in a cloud and the total response time to a given set of queries. First, we optimize the two objectives separately: one criterion is optimized under a bounding constraint on the other one. Our proposal is validated through experiments.