Enabling Compute and Data Sovereignty with Infrastructure-Level Data Spaces

Abstract

Data is a critical asset in today's world. When multiple actors are involved, ensuring that data is accessible only to authorized parties and protecting it against theft present significant challenges. A potential solution to these issues is creating data spaces that interconnect clusters managed by different actors. The latter can securely exchange data under specific constraints and terminate connections when needed. This paper aims to show how infrastructure-level data spaces, which support both access to data, and processing of it, can facilitate secure data exchange and limit data theft. Furthermore, we investigate how data sovereignty can be maintained through cluster data exchange, which is crucial in an era where data is increasingly regulated and controlled. Additionally, we explore how offloading applications from the data consumer into the data producer cluster can match data gravity patterns, improving overall system efficiency. Finally, this paper presents the potential integration of the proposed solution within the framework of IDSA and Gaia-X, serving as promising option for implementing their proposed functionalities.