Scaling Stateful Services: Strategies for Consistent Hashing and Rebalancing?

Hey everyone, I'm thinking about how to best scale stateful services, especially when dealing with data partitioning and ensuring consistency during rebalancing. My current focus is on systems where we need to maintain some form of session state or cached data across multiple instances. I'm leaning towards consistent hashing, but I'm curious about the different algorithms and practical considerations for rebalancing when nodes are added or removed. What are your experiences with specific consistent hashing implementations, and how do you handle data migration and eventual consistency during scaling events without major service interruptions?