This project is developing techniques that can improve the efficiency of data management software (e.g., database and filesystem) by leveraging new data storage hardware with built-in compression. These are disks that automatically compress data when it is written to the disk and decompresses data when it is read. This makes the disk appear larger than it really is because of the compression. For the first time, such new storage hardware decouples the logical storage space utilization efficiency from the physical storage space utilization efficiency, which creates a large but completely unexplored space for simplifying the design and implementation of data management software. This project involves developing a set of techniques to explore this new design space for simplifying mainstream data management software. In order to demonstrate and study the involved design trade-offs, this project will implement prototype key-value store and caching software, and enhance existing open-source filesystems and relational databases.
By developing solutions to largely simplify data management software, this project could significantly contribute to improving the efficiency and cost effectiveness of future data-intensive computing infrastructure. This work can inspire broader research to fully explore the potential of leveraging new storage hardware with built-in computation capability to innovate computing system software stack. The developed source codes and experimental results will be available online for public and wide-range usage, benefiting both industrial and academic researchers. The research training to both undergraduate and graduate students will address the concerns of lacking researchers and engineers with cross-hardware/software expertise in IT and computer industries.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
