I know that SQL Server 2005 does some caching of the execution plan, but will that be enough to make the difference between the clocks between the same query that is executed twice? The first time it takes 3 hours the next time it takes 1 min? Is it possible?
SQL Server will not just cache the execution plan as well as the data
if you do
SET STATISTICS IO ON
, , - , - . , ,
3 , - /, ..
!, , ().
, :
: 3 , 1 , , , ., ..., , (, update/insert/delete, , ), SELECT , ( /).
, :
( SQL-Server):
dbcc freeproccache go dbcc dropcleanbuffers go
, ( " " ) , , . , .
( "" ) , , , .
, 3 , , .
, (, - ).
. , , (, ), .
, : INSERT INTO Table (Field) VALUES (1). , , . 750 10%, 75 , SQL Server , 30 . .
INSERT INTO Table (Field) VALUES (1)
The point here is not that during the execution you had or did not have a growth event. The fact is that if the request lasts 3 hours, you need to understand why. Activity Monitor is a great start. Reading white paper, such as "Waiting for Performance," would be even better.
Source: https://habr.com/ru/post/1734572/More articles:Is box-shadow affect element size? - csslog4net custom log objects and application - c #Removing file to trash in Windows x64 in C # - c #выполнить процесс удаленно с помощью System.Diagnostics.Process - processCan I find out if the JFrame is fully visible? - javaiPhone and iPad - including the class will increase the binary file? - iphoneJtable content will not be updated after adding data as a parameter - javaFlex builder3 does not create html wrapper when targeting flex 4 sdk - flexWhat type of data will I use if I want to save a file with a user using PHP and MySQL? - typesAdd Column to PdfPTable, iTextSharp - c #All Articles