|
|
Posted by Erland Sommarskog on 04/16/06 19:52
Serge Rielau (srielau@ca.ibm.com) writes:
> Lastly I respectfully disagree with the subsequent note in this thread
> that over specification of columns in a view has no performance impact.
I think Peter's question actually was whether there was a performance
impact of saying "SELECT *" instead of listing all columns in the table
explicitly, and for all my dislike for SELECT * in production code, I
can't think of any reason why SELECT * should be any more expensive in
SQL Server.
I completely agree with that best is to include exactly those columns
for which there is an actual need.
One problem I often fight at work is that I want to drop a column, or
drastically change the meaning of it. I suspect that it is not really
in use, but still I find a bunch of stored procedures that return this
column in a result set. But I can also see that these procedures aim
at returning the universe, so I have no idea whether the value is used
for something. (Usually, I end up dropping the column anyway.)
--
Erland Sommarskog, SQL Server MVP, esquel@sommarskog.se
Books Online for SQL Server 2005 at
http://www.microsoft.com/technet/prodtechnol/sql/2005/downloads/books.mspx
Books Online for SQL Server 2000 at
http://www.microsoft.com/sql/prodinfo/previousversions/books.mspx
Navigation:
[Reply to this message]
|