Your question is Partitioning Training Metrics Time Series. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You're designing storage for model training metrics and want query performance to stay predictable as history grows. The data is append-heavy, time-based, and used for both recent debugging and longer-term trend analysis.
How do you approach partitioning and indexing strategies for time-series data related to model training metrics?