Fact : Teradata loads the data into the table in parallel.
I have created a table EMP with UPI on Emp id column.If we are loading some 1million rows into this table Emp which happens in parallel way.
Suppose if there are duplicate Emp id's coming at the same time(i.e with same timestamp) then both rows should be inserted right
,but what will happen is it will insert one row and throw the error as Uniqeness constraint violation for the other row.My question is that how it is able to check UNIQUENESS constraint on the Empid column if the data is being loading into the table in parallel way.
Please shout at me if you need more info but please clarify...........
The answer depends on what method/tool you are using to load the data. In general though, what it means is that all AMPs are working on the process of loding simultaneously. Rows with the same PI will be all be handled by the same AMP so that AMP can check uniqueness.
Considering the above table definition,suppose if two rows with same UPI having the same timestamp goes to the AMP then which row will be inserted into the table.