September 17, 2005 at 6:04 am
What the h3ck are you testing then if you're not ommiting some columns in one view that you are selecting in the other???
September 19, 2005 at 1:31 am
The size is varchar(100).
I am omitting some columns, ofcourse.
Heres the picture
I have two tables
T1 and T3, having similar structures.
fname varchar(100)
lname varchar(100)
srno int (identity)
In query I am omitting column T1LNAME and T3FNAME. Its the same as in the defn for view [t1t3limit].
September 19, 2005 at 7:15 am
So you're querying from 2 different tables???
September 19, 2005 at 7:25 am
Thats right.
CREATE view t1t3
as
select t1.srno as T1SRNO,t3.srno as T3SRNO,t1.fname as T1FNAME, t3.fname as T3Fname, t1.lname as T1Lname,t3.lname as T3Lname
from t1 inner join t3 on t1.srno=t3.srno
CREATE view t1t3limit
as
select t1.srno as T1SRNO,t3.srno as T3SRNO,t1.fname as T1FNAME, t3.fname as T3Fname
from t1 inner join t3 on t1.srno=t3.srno
=====================================================
The select statement is
select T1SRNO, T3SRNO,T1FNAME, T3Fname from t1t3
select T1SRNO, T3SRNO,T1FNAME, T3Fname from t1t3limit
When I execute the above query I get the variations in execution times.
September 19, 2005 at 7:42 am
How about you use 1 view.
Then select all fields from that view, then select all fields but 2-3 from that same view???
Also make sure that the execution plans don't change between each selects.
September 19, 2005 at 7:52 am
If I do that, I get a variance of 5 secs. That is, if i select all the fields from t1t3 and if i select only 4 fields. But thats bound to happen. I want to know is is how does SQL Server query optimiser go about selecting the optimal execution plans for my scenario.
BTW, thanks for your time and responses RGR'us. I appreciate that.
September 19, 2005 at 8:05 am
HTH.
Viewing 7 posts - 16 through 21 (of 21 total)
You must be logged in to reply to this topic. Login to reply