Suppose I have a table in which I have a column which is multi-valued, these values are primary key of another table, now I could normalize this by making it single-valued column and repeating the same row for each value that occurs to be in the column, but I guess this is so redundant, note that this column can have up to 100 values(which are foreign key derived from another table). Now if I make it single-valued column then obviously my rows will be multiplied by the number of keys I take from the foreign table, what is the best practice to normalize this kind of scenario in the most optimized way?
Table-1( pk, column-1, column-2)
Table-2( pk, column-1, column-2, FK(Table-1))
Here my FK can relate to more than one item of table-1
