you are correct. Basically below are conditions which can help you
1. Table is set, we create UPI on it. this way teradata checks for duplicacy of PI values and rejects duplicates , no need to scan entire row and all conditions are satisfied.
2. table is multiset and UPI, here if we just check that no duplciates for PI value, we satisfy all conditions. so no need to scan complete row.
3. table is set and NUPI on it, then complete row needs to be scanned to avoid duplicate rows as PI can have duplicates and not bothered with
4. table is multiset and NUPI, no checks need to be done as everythign is good.
I hope that I helped little bit.
↧