Hasso-Plattner-Institut
Prof. Dr. h.c. mult. Hasso Plattner
 

Autonomous Data Management

The trend of enterprises storing more and more data in cloud-based database systems requires providers to focus on self-optimization and cost-efficiency to stay competitive. We investigate these topics in two connected lines of research.

First, we work on in-memory data management concepts. With our in-memory research database Hyrise, we study how high-performance database systems can optimize themselves with the eventual goal of an autonomous operation without the need for human interaction. To advance in this direction, we are working on various topics, such as data tiering on modern storage technologies, data encoding, index selection, and data dependencies.

The second line of research explores a database system architecture that exploits modern, serverless cloud infrastructure such as Function-as-a-Service (FaaS) platforms to achieve cost-efficiency. With our Skyrise database system, we target the current sweet spot of interactive in-situ analytics on cold data in cloud storage. Skyrise incorporates a number of components from Hyrise and introduces new capabilities to compile and execute queries elastically with cloud functions.

Teaching

We offer lectures, seminars and projects on data management for enterprise applications. For HPI Master's students, we also provide a varity of thesis topics.

Publications

Our research has been published in leading database systems conferences (VLDB, EDBT, ICDE, CIDR, and SIGMOD).

Sorry, the requested view was not found.

The technical reason is: No template was found. View could not be resolved for action "download" in class "AcademicPuma\ExtBibsonomyCsl\Controller\DocumentController".