Serverless vs Containerized Microservices: A Hybrid Cloud Deployment Strategy

Abstract

Use of hybrid cloud deployment strategies that apply serverless computing and containerized microservices are rising in cases that need fast scalability in addition to fine grained management. Serverless architectures do better in supports bursty, event-driven workloads and have lower operational overheads whereas containerized microservices can offer predictable performance and orchestrating flexibility of sustained services. Edge-optimized distributions of Kubernetes carry these paradigms to low-latency, locality-sensitive settings. Stitching workflows in such a way that they cut across these models of execution achieves an ideal placement of workload distributed across public clouds, on-prem, and edge nodes. Nonetheless, performance inconsistency, cold-start start, cost-utilization tradeoffs, and in reproducibility are still in play. The overall trends of such studies concerning the architectural basis, the performance properties, the integration approaches, and optimization models are reviewed with the conclusion of open research questions and guidelines of best practice in adopting hybrid cloud performance.