The table has a PRIMARY KEY and a non-unique INDEX on a BIGINT column. My question is, how can we run a query on this INDEXED column and figure out which PRIMARY KEYs have the most matches. As of right now, this is pretty much as far as I can get it:
indexed_column = some_numeric_value
I know there are tiny ways to improve this, but I am looking to go from several seconds to getting data within as fast as 100-200 ms. Any ideas? Am interested in hearing any algos, tuning tips, or even database specifically designed for this even if it is not MySQL.