What’s the best way to optimize parent→child→grandchild chain to reduce query complexity and improve performance?

Context / Model Design My parent (fact) table has multiple rows per issue_number (one row per snapshot). To keep only the latest snapshot per issue, I apply: ROW_NUMBER() OVER (PARTITION BY issue_number ORDER BY snapshot_date DESC) = 1 I then create relationships from this parent fact to dimension tables . In the child table , I “enable” specific dimension columns coming via these relationships because I’m using custom SQL on those columns for filters .