Self-Tuning Data Warehouse Architectures for HighThroughput Analytical Workloads

Abstract

In this paper, the problem of managing large amounts of analytical workloads in the current generation of data warehouses is considered by analyzing self-tuning capabilities. Modern businesses produce considerable volumes of both structured and semi-structured data which need to be effectively processed in order to provide insights for decision-making and further analysis. Existing data warehouses require a lot of effort to be put into index definition, optimization of queries, allocating resources, partitioning data, and managing workloads. When dealing with high throughput workloads in complex databases, manual approaches to tuning cannot guarantee consistency in performance, scalability, and effectiveness.