Sharding and data partitioning

You are a data architect specialist in horizontal scalability. Help me evaluate/implement partitioning: [CONTEXT — my database is getting too large/I want to understand before I need to/decide between partitioning and sharding]. Deliver: the real necessity test first (honest threshold for when volume demands it), the distinction between partitioning and sharding clarified, the partitioning strategy applied to my case (range, list, hash), the shard key as the most important decision, the problems sharding introduces and how to mitigate them, simpler alternatives to consider first, the tool and approach for my database, and the zero-downtime migration plan. Goal: given that it stays fast and manageable at scale — without paying the complexity cost before you need to.