Say you had a column and the clustered index is on that column and that
column alone, now, the values inserted into that column are always inserted
at the end of the table (for want of a better description), for example, the
column has the IDENTITY property.
Also, there are no deletes to the table.
Now, the index can never get fragmented right?
Logically - no; however - the rows may have been inserted fragmented in the
first place because of the availability and location of free extents in the
database. Basically, your table could be mixed up with other tables which
can effect read aheads.
So, to answer your question - its still worth rebuilding indexes but you
base that decision by looking at fragmentation stats out of dbcc showcontig
or sys.dm_db_index_physical_stats.
Tony.
--
Tony Rogerson, SQL Server MVP
http://sqlblogcasts.com/blogs/tonyrogerson [Ramblings from the field from a SQL consultant]
http://sqlserverfaq.com [UK SQL User Community]
[quoted text, click to view] "Ed" <Ed@discussions.microsoft.com> wrote in message
news:3DE386DD-BF28-4C17-9404-170A107D1B58@microsoft.com...
> Hi,
> If the leaf level of Clustered Index is the actual data page, is that
> necessary to reorganize the clustered index? I undstand the non-leaf and
> root level but should not be at the leaf level. Am I right?
>
> Thanks
>
> Ed