Abhishek Singh: LogPoseDB: Transaction Handoff and Agreement in Edge-Cloud Systems

Abstract: Emerging IoT and edge applications demand fast response times that cannot be achieved by faraway cloud datacenters. This motivates building edge-cloud systems where nodes on the edge can participate in the processing and storage of data. However, building an edge-cloud transaction processing system faces two main challenges: (1) Inefficient transaction execution due to low concurrency arising from long  Round Trip Times between Edge and Cloud, (2)The absence of a dedicated edge nodes infrastructure, and (2) edge nodes may be untrusted.
We propose LogPoseDB, an edge-cloud database that spans both edge and cloud nodes. LogPoseDB aims to overcome the challenges above. LogPoseDB proposes dynamic state detachment, where the state storage and processing is treated as a disjoint resource between the cloud and the edge. LogPoseDB’s transaction processing protocol ensures fast response by avoiding wide-area coordination with the cloud or other faraway edge nodes. This is done by leveraging data locality of detached state and by methods that build on the areas of transaction chopping and commutativity.
LogPoseDB does not require any dedicated edge infrastructure. Rather, clients may utilize their edge nodes—if desired—to perform the processing and storage of their data while they need it. (Other clients can still process their data on cloud nodes.)  To address the trust challenges, we propose a byzantine fault-tolerant (BFT) protocol that targets edge nodes. LogPoseDB’s BFT replication protocol proposes the principle of  remote lazy trust that enables efficient BFT edge coordination by utilizing a remote trusted node asynchronously.

Bio: Abhishek is a PhD Candidate supervised by Prof. Faisal Nawab. His research includes building data management and transaction processing systems for the emerging Edge-Cloud infrastructure.