Dynamic Power-Aware Disk Storage Management in Database Servers

Energy consumption has become a first-class optimization goal in design and implementation of data-intensive computing systems. This is particularly true in the design of database management system (DBMS), which was found to be the major consumer of energy in the software stack of modern data centers. Among all database components, the storage system is the most power-hungry element. In this paper, we present our research on designing a power-aware data storage system. To tackle the limitations of the previous work, we introduce a DPM optimization model to minimize power consumption of the disk-based storage system while satisfying given performance requirements. It dynamically determines the state of disks and plans for inter-disk fragment migration to achieve desirable balance between power consumption and query response time. We evaluate our proposed idea by running simulations using several synthetic workloads based on popular TPC benchmarks.