Reproducibility and Testing of Computational Storage

Supervisors: Niclas Hedam, Philippe Bonnet
Semester: Fall 2023
Tags: Open Source, Testing, Computational Storage, Hardware, FPGA

The emergence of computational storage platforms like Delilah has transformed the data storage landscape, enabling new computing paradigms and facilitating data-intensive applications. Delilah is a cutting-edge computational storage platform developed by the IT University of Copenhagen. It runs on the Daisy OpenSSD and exposes an asynchronous computational storage protocol to the host, facilitated by eBPF, which orchestrates I/Os to the underlying storage mediums.

However, Delilah’s quality assurance and regression testing in its specialised environment pose a significant challenge, necessitating cumbersome manual interactions. Given that Delilah is currently being used in active research through the DAPHNE project and through collaborations with the Technical University of Dresden, further work on its development has the potential for significant impact.

This research project aims to investigate and propose efficient and reliable methods for testing Delilah. As a starting point, the student can reorganise Delilah’s source code to enable unit testing or implement mocking mechanisms through feature testing to detect and verify operations before they leave Delilah.

This research project presents a unique opportunity for students to engage in cutting-edge research in computational storage and contribute to the development of Delilah. This project is intended to be a Research Project to be followed up with a Master’s Thesis. The research findings are expected to contribute to the advancement of computational storage platforms and have significant implications for the broader field of data-intensive systems.