what is the best way to get distinct records from the table containing 500 mln records? table structure is id1, id2, id3, upload_date. sample data
id1 id2 id3 upload_date
1 ab cd 20240525,
2 null gh 20240331,
2 null gh 20240228