in my project, the Sql server DB contains lot of customer information into different tables. The information about the customer is huge but the actual number of customer are less comparatively hence I want to partition the DB on the basis of CustomerID so that I can improvement performance. For example, the main table which store customer information is tblCustomer. it's Primary key i.e. CustomerID which is used in all other tables as foreign key and there many tables which are using CustomerID column as FK.
I want partition to do partitioning on CustomerID column.
Can I partition on database Or tables only ? Should I use non-clustered index OR partitioning ? If partitioning then Do I have to partition all the tables ? Do i have to partition on main table i.e. CustomerID ?
Please suggest