I need to create table with 10 columns, Column1 is the unique in entire table and Column2 will have 10 or 15 values. But we query the table based on Column2 value.
In this case on which column we need to create PI which helps quick read and also should help while writing as well. This table would be very huge more than 100 millions in future
2. If we use SET and Non-Unique PI, does that table allows, does it checks for Duplicates while loading. Does the Duplicate means whole row or only column on which PI is same?
Could someone help me on this?
Thanks in advance.
How about Partitioned Primary Index
primary index (col1) partition by (col2);
Any condition on col2 would limit the scans of data to the relevant rows.