CloudIX · CloudIX: Cloud-based Indexing and Query Processing

The advent of cloud computing comprises a new paradigm that entails a large number of low-end processors that perform parallel processing of (usually data-intensive and resource-demanding) computing jobs. This is particularly appealing for a wide variety of data-intensive applications, including scientific data management, analysis and mining of complex data, and internet-scale service provisioning. Nevertheless, processing complex queries in the cloud is still in a primitive stage, largely relying on the concept of map-reduce, without inherent indexing mechanisms that can boost the performance of query processing. In contrast, database research prides itself on scalable techniques for advanced query processing.CloudIX intends to transfer knowledge from data management technology to the cloud infrastructure, in order to support advanced query processing. To this end, CloudIX aims to propose a generic, distributed, dynamic and adaptive indexing framework for handling multidimensional data in the cloud. Capitalizing on this indexing infrastructure, CloudIX will present novel efficient algorithms for processing advanced query types, such as range, nearest-neighbor, top-k, rank joins, and skyline queries. Furthermore, CloudIX will propose new economical cost models that need to be tightly integrated in the process of query optimization.The research results of CloudIX are expected to act as a catalyst that attracts applications that rely on advanced query processing to be moved to the cloud. As a showcase scenario, two applications will be deployed on top of CloudIX, one related with scientific data management, to demonstrate the feasibility and efficiency of the proposed framework.