Hybrid Transactional/Analytical Graph Processing in Modern Memory Hierarchies (#TAG) (SA782/28-2)
PI: Kai-Uwe Sattler (TU Ilmenau)
Project collaborator: Philipp Götze
Project website: to be announced
Today’s enterprise computing architectures are characterized by a complex memory hierarchy: the different application requirements in terms of latency, bandwidth, persistence, and access pattern as well as the characteristics of available memory and storage technology require combining different technologies.
Building highly efficient data management and analytics solutions which meet the challenges of modern applications requires to utilize this memory hierarchy, e.g. by caching strategies, taking the specific characteristics of a given technology into account, and keeping data objects in the optimal level. In this project, we plan to exploit modern memory hierarchies to support Hybrid transactional/analytical processing (HTAP) on graph data.
The project has three main objectives: First, we plan to design, develop, and evaluate data structures and query operations for graph data in persistent memory. The goal is to simultaneously support transactional changes, navigational queries, and graph analysis. A second goal is to support analyses of graph data on hardware-based accelerators with dedicated memory (e.g. GPUs) by efficient data transfer and consistency mechanisms. Third, we consider time-travel queries by swapping temporal differential snapshots of graph data also to block-oriented external storage and including them in query processing.