Within the Uklon service product line, we consistently employ a widely accepted and easily understandable programming paradigm — the stateless approach. However, there exist various business scenarios where this approach may fall short when it comes to resource utilization, efficiency, and ongoing maintenance. Furthermore, its predictability in handling peak loads can also be questioned. In this presentation, we will explore the following key areas: Scenarios where the stateless approach demonstrates suboptimal performance; Scaling stateful services through effective partitioning and replication strategies; Developing strategies for comprehensive disaster recovery, including Recovery Time Objective and Recovery Point Objective; The evolutionary path from actor-based architectures to the more inherent classic stateful architecture.