Hello everyone,
I have a couple of tables (SQL 2005) where I have created a clustered index on the PK that is a uniqueidentifier.
The performance of a query that joins these two tables has indeed improved (not exactly sure why, since I am not specifying a range, sorting or grouping).
I was under the impression that a clustered index on a unique identifier column, due the random nature of the GUIDs, will lead to index fragmentation. But I am not seeing that effect (at least not yet).
Is a clustered index on a unique identifier OK?
Or will it invariably lead to performance degradation?
What are your thoughts on this topic?
Thank you,
Sergio.