Consider table partitioning for per-block tables

[ptrus]

Most queries primarily focus on recent block data. In nearly all endpoints that return lists of transactions, blocks, events, etc., we use ORDER BY block DESC. Additionally, per-block tables such as blocks, transactions, and events have grown significantly.

The database would likely benefit greatly from table partitioning (e.g., partitioning by every 1 million blocks or similar). We should conduct tests and consider implementing this in the future to improve performance and optimize database load.