Applicability of Lightweight Groups to Fog Computing Systems
Diogo Lima, Hugo Miranda · 2021
The tradeoff between scalability and consistency has been addressed by deploying distributed applications in the cloud. However, cloud-based applications penalize performance with the unavoidable latency and jitter resulting from the geographical distance between clients and application servers. Fog Computing architectures mitigate this impact by deploying state fragments and computing power on surrogate servers at the network edge. To improve performance, one must deploy each fragment of the application state at the most convenient surrogate and maintain an efficient coordination when manipulating state stored in multiple locations. This paper formalizes a two-tier view synchronous group communication model for Fog Computing to improve performance of large scale mobile distributed applications and evaluates it using data derived from a real dataset.