Features of Developing Front-end Technologies in Cloud Environments: Client–Server Architecture and Temporal Modeling

The paper examines the features of front-end technology formation in cloud environments based on the client–server model and provides mathematical justification for applications of the Smart City type. A generalized latency model of user requests is proposed, in which the components correspond to network delays, server processing, data access, and serialization/rendering costs in the browser. For multi- endpoint REST interfaces, a weighted evaluation is introduced that accounts for the frequency and criticality of calls, while for aggregated GraphQL queries, a parallel model is proposed that captures the impact of the “slowest” dependency. It is shown how the parameters of user request latency can be reduced by means of cloud technologies, including CDN/Edge networks and HTTP/3, API Gateway/BFF with partial delivery, caching and incremental queries, as well as progressive/streaming rendering and Web Workers on the client side. For capacity planning, the M/M/N queuing system (Erlang-C model) is applied to estimate expected waiting times, and SLO benchmarks are established for urban scenarios. Using the example of a dashboard integrating transportation, air quality, and power grid data, it is demonstrated that GraphQL reduces the number of requests and stabilizes latency compared to multiple REST calls, whereas REST remains effective for simple, well-cached resources. The results provide practical recommendations for designing cloud-based front-end architectures with predictable perfor- mance and manageable operational costs.