I have large set of data (almost 30 millions) in mongodb collection which has _id(mongodb default id) creation_date (creation date of that document) and symbol (predefined only 50 symbols and symbols are constant. Does not varies.). I am searching on two field i.e. creation_date and symbol. Let say I want last 30 days data for all 50 coins.

NOTE: Data in collection are equally divided with 50 symbols.

Question: What is best approach to at least make the searching faster.

What I know which could be potential solution: Sharding (on collection) on date range. May be 1 month range

Please give your approach. Because I am stuck. Actually I haven't tried sharding yet. Would it be good solution.