Home/ Questions/Q 93749
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-05-10T23:23:36+00:00

I’m using MS SQL Server. When I define the database schema I define a

  • 0

I’m using MS SQL Server.

When I define the database schema I define a (non-materialized) view, which includes many fields, for example as follows (where ‘Topic’ is the name of a table, and the view is a self-join on the Topic table):

CREATE VIEW View_Topic_Ancestor AS SELECT     Subordinate.Id AS Subordinate_Id,     Subordinate.Folder_Id AS Subordinate_Folder_Id,     Subordinate.topicTitle AS Subordinate_topicTitle,     Subordinate.topicXhtml AS Subordinate_topicXhtml,     Subordinate.crossLinked AS Subordinate_crossLinked,     Superior.Id AS Superior_Id,     Superior.topicTitle AS Superior_topicTitle,     Superior.topicXhtml AS Superior_topicXhtml,     Superior.crossLinked AS Superior_crossLinked FROM Topic AS Subordinate LEFT OUTER JOIN Topic AS Superior ON Superior.Folder_Id = Subordinate.Folder_Id AND  Superior.LeftValue = (SELECT MAX(Ancestor.LeftValue)     FROM Topic AS Ancestor     WHERE Subordinate.LeftValue > Ancestor.LeftValue     AND Subordinate.LeftValue < Ancestor.RightValue     AND Subordinate.Folder_Id = Ancestor.Folder_Id)

Later (at run-time) I use this view in a select statement, like this:

SELECT     T.Id AS Shared_Id,     V.Superior_Id,     V.Superior_topicTitle,     V.Subordinate_Id,     V.Subordinate_Folder_Id,     V.Subordinate_topicXhtml FROM Topic AS T, View_Topic_Ancestor AS V WHERE Folder_Id='e2eb2b68-738d-49ad-9787-a1e655b7973f' AND T.crossLinked = V.Subordinate_Id

This SELECT statement doesn’t reference (doesn’t select) many of the fields which are in the view: for example, it selects the Subordinate_topicXhtml field but it doesn’t select the Superior_topicXhtml field.

My questions are:

1) Do the fields which are defined in the view, but which are not referenced in the run-time selection from the view, have much effect on performance? Assume if you will that the Superior_topicXhtml field contains a lot of data (is a very long string).

2) How can I verify the answer to this myself? Is testing (measuring ellapsed time with a stop-watch) the only way, or is it possible to obtain an answer based on theory? I am using ‘Microsoft Server SQL Management Studio’ for Microsoft SQL Server 2008, with SQL Express. I see how to obtain (but haven’t learned how to interpret) the ‘estimated execution plan’ for this query, but this shows only what indexes and loops are happening, not whether data is being retrieved from unreferenced fields.

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. The Query optimizer creates a query plan by combining the sql which defines the view with the sql you construct to ‘select’ from the view into a single sql query. If, in this combined, optimized sql, any specific attribute (column) is not necessary, it will not be used.. specifically, if the final select does not require it, it will not be sent over the wire. Also, if it is not needed by any Where clause, or Order By or anything else within the optimized sql, and all the attributes that ARE needed happen to be in an index, then that query can read only the index, and not touch the table at all… even thought he view references a column not present in the index… You can see this by looking at the query execution plan for the query.

    • 0