K
kevin.c
This is somewhat of an odd situation...but to make a long story short:
I built a test web application which inserts X number of records into a
table. It does this via a stored proc, shown here:
CREATE PROCEDURE dbo.spInsertTestData
(
@TestID uniqueidentifier,
@TextData nvarchar(100)
)
AS
SET NOCOUNT ON
INSERT INTO tblTest
(TestID,TextData)
VALUES
(@TestID,@TextData)
-----
fairly simple....Im using two different methods to do the inserts,
creating a SQLConnection and SQLCommand object myself, or letting DAAB
do it.
On my development machine the results of 300 inserts are as follows.
Bare in mind the distance between the development web server and the
SQL server is rather large and these results are acceptable:
Non-Daab: 22.94, 21.65, 21.56 (secs)
DAAB 25.73, 21.91, 21.67 (secs)
Transfering the application to the production servers (where the
distance between the servers is about 6inchs on a gigabit connection):
Non-Daab: .21,.20,.23
DAAB: Timeout......
Dropping the record count to 5 records on the production machine I get
results of
Non-Daab 0,.1,.1
Daab, 4.87,5.63,4.83
DAAB is basicly taking 1 second per insert on the production machine,
but runs at the same speed as the non-daab methods when ran from a
development box. This holds true to running against a test sql server
that is local to the dev machine. Results are all between 0-2 seconds
for doing 300 inserts on both tests. It boils down to something being
different on the production machine that is causing DAAB to run
incredibly slow.
Any ideas?? We are really out of options, lol. The production machines
are faster servers than the test box, yet it runs far far slower. All
of our testing says its something that the DAAB dlls are trying to do
that is very cumbersome to that production machine. However, we cant
see any differences in the setup (between production and development
machines). What is/does daab do under the surface that could be
causing this. We've taken a quick glance at the daab source code, but
cant find anything that sticks out like a sore thumb.
Thank you for your input
I built a test web application which inserts X number of records into a
table. It does this via a stored proc, shown here:
CREATE PROCEDURE dbo.spInsertTestData
(
@TestID uniqueidentifier,
@TextData nvarchar(100)
)
AS
SET NOCOUNT ON
INSERT INTO tblTest
(TestID,TextData)
VALUES
(@TestID,@TextData)
-----
fairly simple....Im using two different methods to do the inserts,
creating a SQLConnection and SQLCommand object myself, or letting DAAB
do it.
On my development machine the results of 300 inserts are as follows.
Bare in mind the distance between the development web server and the
SQL server is rather large and these results are acceptable:
Non-Daab: 22.94, 21.65, 21.56 (secs)
DAAB 25.73, 21.91, 21.67 (secs)
Transfering the application to the production servers (where the
distance between the servers is about 6inchs on a gigabit connection):
Non-Daab: .21,.20,.23
DAAB: Timeout......
Dropping the record count to 5 records on the production machine I get
results of
Non-Daab 0,.1,.1
Daab, 4.87,5.63,4.83
DAAB is basicly taking 1 second per insert on the production machine,
but runs at the same speed as the non-daab methods when ran from a
development box. This holds true to running against a test sql server
that is local to the dev machine. Results are all between 0-2 seconds
for doing 300 inserts on both tests. It boils down to something being
different on the production machine that is causing DAAB to run
incredibly slow.
Any ideas?? We are really out of options, lol. The production machines
are faster servers than the test box, yet it runs far far slower. All
of our testing says its something that the DAAB dlls are trying to do
that is very cumbersome to that production machine. However, we cant
see any differences in the setup (between production and development
machines). What is/does daab do under the surface that could be
causing this. We've taken a quick glance at the daab source code, but
cant find anything that sticks out like a sore thumb.
Thank you for your input