Evolutionary multiobjective query workload optimization of Cloud data warehouses

Area of Science:

• Computer Science
• Database Systems
• Cloud Computing

Background:

• Cloud databases present challenges for query optimizers balancing response time and monetary cost.
• The elasticity of cloud virtual resources offers opportunities for optimizing query execution.

Purpose of the Study:

• To develop and evaluate novel multiobjective optimization algorithms for distributed data warehouse query workloads in the cloud.
• To minimize both response time and monetary cost simultaneously.

Main Methods:

• Proposed an exact multiobjective branch-and-bound algorithm.
• Developed a robust multiobjective genetic algorithm.
• Integrated algorithms into a prototype system for evaluation.

Main Results:

• Demonstrated the effectiveness of the proposed algorithms through experiments with varying workloads and resource configurations.
• Identified optimal deployment strategies leveraging cloud virtual resource elasticity.
• Analyzed the trade-offs and performance characteristics of the multiobjective algorithms.

Conclusions:

• The novel approach effectively minimizes response time and monetary cost for cloud data warehouse queries.
• The study provides insights into the advantages and disadvantages of the proposed multiobjective optimization techniques.
• Findings highlight the potential of alternative deployments and algorithmic choices in cloud database optimization.