We witness a shift from centralized to decentralized distributed systems. In a decentralized system data is typically copied (i.e. replicated) to various (potentially malicious) machines. As a result, programmers need to 1) manually synchronize these copies such that they remain consistent, and
2) ensure that sensitive data is not leaked to unauthorized users. The synchronization mechanism determines the system's availability, i.e. whether or not the system accepts reads and writes at all times.

We propose a general-purpose framework for replication which allows programmers to trade-off the consistency and availability of individual objects, and to control the dissemination of data. The overall goal of this research is to design a new programming model that hides the details of data replication but also provides the necessary privacy abstractions. As such, we tackle both the complexity of replication and the trust issues which are inherent to decentralized systems.
Effective start/end date1/01/1931/12/20

    Flemish discipline codes

  • Other computer engineering, information technology and mathematical engineering not elsewhere classified

    Research areas

  • Application development

ID: 43836591