All those unique indexes seem suspicious, are they actually based on access/join paths or do they just implement the logical keys? A NUPI (FIRST_DT ,VISITOR_ID) on T_CUST would allow a direct join without preparartion.
Joining up to 1.300.000.000 rows and aggregating them is a lot of work, can you show the explain?
All those unique indexes seem suspicious, are they actually based on access/join paths or do they just implement the logical keys? A NUPI (FIRST_DT ,VISITOR_ID) on T_CUST would allow a direct join without preparartion.
Joining up to 1.300.000.000 rows and aggregating them is a lot of work, can you show the explain?