We offer the conceptual model of the hybrid geographic information system built by using Kubernetes, Docker containers and cloud computing. Within the study, we conducted stress tests and velocity measurements in virtual environments. Applications deployed in containers show better performance when one needs to simultaneously run the same type of isolated compute nodes. The experimental results clearly demonstrate the advantage of using Kubernetes with Docker containers to build such systems, especially in case of stateless applications. However, since we are talking about stateful applications, it is necessary to solve the data storage problem in order to ensure the system's stable operation. Today there exist several standard approaches for solving this problem. All of them are presented in the article. The article contains our practice-based recommendations on using these approaches in various situations. The article summarizes the theoretical and practical experience gained by testing this technological stack when implementing various web services, from personal sites to corporate portals. As a result of analyzing such approaches, we propose the conceptual GIS hybrid model based on the Kubernetes computing cluster with Docker containers for data processing, whereas external cloud storage is used to store data and backups. In our opinion, the described conceptual model of IT infrastructure can be used to build any web services requiring scalability, system fault tolerance and security.