Flexible Data Placement on SSDs for Database Systems

Supervisors: Pınar Tözün
Semester: Fall 2024
Tags: SSDs, data management systems, modern storage

Traditionally solid-state drives (SSDs) does not give the users the ability to control the data placement on the SSD. This often leads to suboptimal performance and lowers SSD lifetime, since SSDs internally don’t allow in-place updates. The updated disk pages are written elsewhere and the old versions have to be garbage collected. This poses problems if data with different lifetimes and update frequencies often gets written in the same region.

To cope with this issue, there has been several attempts by storage vendors such as open-channel SSDs, streams, zoned-namespaces, etc. The most recent attempt is, though, a new SSD type with support for flexible data placement (FDP). This new class of SSDs allow hints from programs to the SSD to dictate which data should be written where, thereby allowing separation of data from different applications or with different lifetimes.

The goal in this project is to perform a feasibility study for the FDP technology using DuckDB. The first step would be integrating FDP with DuckDB. Then, we would analyze the performance impact of this integration.

If you are interested in data management systems, storage devices, benchmarking, and performance analysis in general, this project would be a great fit for you.

This project would be the most suitable as an MSc project that starts with during Fall 2024 with the research project phase and continues into the thesis. However, if you are interested in this topic for a regular research project or BSc thesis, please reach out still.