Why is the unique argument in the index creation statement?
As I have understood, the non-cluster index should be unique to a bookmark, an indicator in a row, which should be unique to differentiate between non-unique rows;
-Clustered index should be exclusive? Is that right?
So, do I think that any unique index can not be on the cluster table?
- "A compiled index should be unique on one view" [1]
Because "the bottom, or the leaf, the level of the package index is the table's actual data Rows "[1], do I correctly believe that the same effect as the UNIUE on the package index can be achieved by a unique obstacle on a table [possibly all or some part] [2]? >
Then, what would bring unique logic for the index?
Definitions of basic concepts, except illusions [3]
UPDATE:
It is again the same threat - explained some already explained several times already Without that it can be understood that all explanations are based on the changing game.
Please see my subquisan [4] which is actually the re-wording of this same question.
Update 2:
The problem is in ambiguous, inappropriate use of words in definitions or inappropriate contexts. If the index is defined as the (identity and) identification / pointing structure for the actual data, then the non-unique or null index does not make any sense. Bye
Cited:
[1] Create Indexes (Transact-SQL)
[2]
Create Tabs (Transact- SQL)
[3]
Unique index or unique key?
[4]
What is the index and the non-clustered index is exclusive
Comments
Post a Comment