Exploring the Internals of DuckDB: A Course Overview

In a comprehensive 15-week course designed for undergraduate students at the University of Tübingen, I delve into the intricacies of the DuckDB relational database system. This course, crafted with the support of the Database Research Group, provides a structured journey through the selected internals of DuckDB. While the duration doesn't allow for an exhaustive exploration of every component, key topics include query performance, memory management, and vectorized query execution.

The course begins with a welcoming setup and progresses through various critical aspects such as sorting large tables, the ART of indexing, and the development of query execution plans and pipelining. We also cover query rewriting and optimization, ensuring that students gain a robust understanding of these essential elements. Basic SQL skills are a prerequisite, though the course primarily focuses on the core SELECT-FROM-WHERE-GROUP BY-HAVING block, with additional resources available for those needing a refresher on SQL and the tabular data model.

This educational journey is supported by slides and auxiliary materials available on a GitHub repository, providing students with the tools they need to succeed. The course is part of a broader set of resources, including talks and podcasts, that explore the latest developments and future directions for DuckDB, such as the intriguing 'Super-Secret Next Big Thing for DuckDB'.