Viewing 15 posts - 106 through 120 (of 342 total)
I would take a stab at the staging table. We currently run a process that receives over 1 million entries per day. The database is able to keep...
September 3, 2003 at 12:08 pm
Could you use a view?
Table (
Field1,
Field2,
OrigPwd VARCHAR(20)
)
CREATE VIEW v_Table AS
SELECT Field1, Field2, '******' AS OrigPwd....
Guarddata-
September 3, 2003 at 8:25 am
You might try
INSERT INTO <table>
SELECT val1, val2, COUNT(*) FROM <table>
Guarddata-
September 3, 2003 at 8:20 am
I'm sure there are better descriptions of transaction handling, but I hope someone else known the reference. I was fortunate to have a couple of days to test different...
September 3, 2003 at 8:18 am
Sorry Lenard. Perhaps if we understood more of the background we could help find another solution. Sounds like you are pretty adept. Good Luck!
Guarddata-
September 2, 2003 at 5:21 pm
Sorry Lenardd - I left out that part of the response . I was thinking you could use this in place of the function. Is...
September 2, 2003 at 3:50 pm
That is correct - I ran into that problem also a while back (no sp_executesql inside a function). Kind of an obstacle.
Perhaps you can build the SQL like
'SELECT '...
September 2, 2003 at 12:34 pm
You are correct about the Auto-Commit...but that applies to the entire procedure. In order to break it down we use the BEGIN and COMMIT.
Guarddata-
September 2, 2003 at 11:16 am
Actually, lenardd is on the right track with the OUTPUT handling in the sp_executesql. I have been able to use variables nicely with dynamic SQL.
The greater challenge in my...
September 2, 2003 at 9:19 am
Yes - if there is an index that matches the WHERE clause - everyone is happy.
The BEGIN and COMMIT are necessary. Without them the process runs in a single...
September 2, 2003 at 8:51 am
The way to tell if the script actually worked is to run it in Query Analyzer. You should see that the results returned indicate 290391 rows affected once, then...
August 29, 2003 at 11:14 am
buktseat, Interesting request to have all the output in a single field. I assume this is for display.
Ordinarily, I would have done
SELECT CaseID,
SUM( CASE StatisID WHEN 4...
August 29, 2003 at 8:29 am
Mind still not in gear this morning... JPipes - why do you say two indexes would be needed?
We have even had good success by using ROWCOUNT:
SET ROWCOUNT 10
SELECT * FROM...
August 28, 2003 at 8:40 am
One more thing to remember... when you use TOP #....ORDER BY the system must produce the entire result set anyway. Performance improves significantly when you can use an index...
August 28, 2003 at 8:27 am
In previous versions of MS-SQL, cursors were such poor performers than almost any other method would have been preferred. The last set of tests we ran on SQL 2000...
August 26, 2003 at 8:20 am
Viewing 15 posts - 106 through 120 (of 342 total)