I want to know what is the difference between query and view in terms of performance. And if the view is expensive, what else can you do to improve performance besides the request?
Representations and special requests in the simple case are almost identical in terms of performance. So much so that when you program with a view, you should think about it as if the text of the view definition was cut out and pasted into your parent request.
HLGEM states in its response that some editions of SQL Server allow you to “index” views — in this case, behind the scenes, SQL Server supports the same structures that underlie the table, making the indexed view and the table very similar in terms of performance.
In SQL Server, although you can usually easily expose views without running into performance issues, this can make understanding and debugging difficult.
, SQL Server , Enterprise. , , , . , , . , , . (, , .)
, . , .
( , ), . , , , , , , , . , , , , , .
, : , . , .
, . , if ( , , ) , , where.
, , - . , , , 4,459,203 , 10, . , , ; , , - .
SQL Server , . , , - , . , , , , . . . , concurrency...
, , , , , , . , , SQL Server Enterprise , .
, , . , .
...
, "-" . . , , , , , , , . .
, . / , , , . , ( ) .
, , , . , , ?
- , , FROM, LEFT OUTER JOIN . , , , "" - , . "" - . - , .. ( SQL-92 - , .)
If you mean network performance, then working with a local cache (as with ADO.Net DataSets) will reduce network traffic, but may cause blocking problems. Just a thought.
The view is still a request, it simply abstracts some parts of it so that your requests can be simplified (if they do such things) and maximize reuse.