Advanced Partitioning Strategy for Time-Series Data

Create a sophisticated partitioning strategy for a high-volume time-series database handling 10 million writes per hour. Develop a solution that implements rolling time-based partitions, automatic partition pruning, and efficient data archiving mechanisms. Address performance challenges including read/write optimization, storage management, and seamless historical data retrieval without compromising query speed.