Greedy Nominator Heuristic: Virtual function placement on fog resources

Fog computing is an intermediate infrastructure between edge devices (e.g., Internet of Things) and cloud systems that used to reduce latency in real-time applications. An application can be composed a collection virtual functions, which dependency constraints captured service function chain (SFC). Virtual functions within SFC executed at different geo-distributed locations. However, are prone failure often do not complete deadline. This results reallocation other nodes the infrastructure; causing delays, potential data loss during migration, increased costs. We proposed Greedy Nominator Heuristic (GNH) address these issues. GNH based on redundant deployment tracking functions. places replicas each multiple locations—taking account expected completion time, risk, cost. make use MapReduce-based mechanism, where Mappers find suitable locations parallel, Reducer then ranks Our show reduces by up 68%, more cost effective than approaches rely state-of-the-art optimization algorithms allocate replicas.