Home/ Questions/Q 578285
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-05-13T14:14:38+00:00

Excerpt from: http://www.sqlservercentral.com/articles/Indexing/68563/ The width of the clustering key does not, however, only affect

  • 0

Excerpt from: http://www.sqlservercentral.com/articles/Indexing/68563/

The width of the clustering key does
not, however, only affect the
clustered index. The clustering key,
being the rows’ address, is located in
every single nonclustered index. Hence
a wide clustering key increases the
size of all nonclustered indexes,
reducing their efficiency as well

Why should the size of the pointer to the datapage in the non-clustered index increase if the sizeof(int *) is the same as the sizeof(char *). Or is there some other addressing mechanism I am not aware of?

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    The pointer does not point to the data page: it points to the clustered index. Subtle, but different to where there is no clustered index where each NC index entry points to the Row ID (RID) in the data pages.

    So if you have char(10) as your key, each NC index entry has a 10 byte pointer. If you have integer, then it’s 4 bytes.

    For char vs int, there are other considerations: collation at least (which determines case, accent, kana and width sensitivity).

    And you string of 4 bytes only applies fot char(4). nchar(4) is 8 bytes, varchar(4) is 2 to 6 bytes (2 byte length), nvarchar(4) is 2 to 10 bytes.

    • 0