I have PI on (COL1,COL2,COL3) of a table. If I use col1 and col2 in WHERE clause
while fetching the data from the table,then is the tables data gets redistributed
according to (col1,col2) rowhash combination to display result set or is the teradata
clever enough to find out result set from already rowhashed value of (col1,col2,col3)
Not 100% sure that I understand your question.
But assume you have a query
where col1 = X and col2 = Y
where yourTable has a PI on (col1,col2,col3).
You will see in the explain a full table scan - no PI access - but no further redistribution of the data.