I would like to ask you two questions about indexing: 1) Since a primary index, or clustering index, stores the tuples of a relation in the primary index itself (but primary index might also be separated from the file containing the tuples), how can we implement this kind of indexes? 2) When we associate to a file a primary index, the file itself must be sequentially ordered. Is true that a primary index (not separated from file) is always an index-sequential file? Thank you.



Browse other questions tagged or ask your own question.