How would you optimize a query that joins a high-volume market tick table with a slow-changing metadata table to minimize execution time?
