Freshness-aware Data Management in a Polystore System (Master Thesis, Finished)
Author
Description
The freshness of data is an important attribute in data management. It reflects "how up-to-date" the data is. The required level of freshness depends on the application. For some applications and queries (e.g. of analytical nature) also slightly outdated data is often sufficient.
Polypheny-DB currently keeps all placements (replicas) of a table consistent. Delaying updates on some of the placements can allow for much higher performance in scenarios where also slightly outdated data is acceptable.
The goal of this project is to integrate the notion of freshness into Polypheny-DB. This includes an adaptive routing model which optimizes the routing of DML queries. Furthermore, it is required to develop approaches for making the required freshness specifiable. An important part of the project is to ensure the overall consistency of the data.
Start / End Dates
2021/11/08 - 2022/05/07