Slow processing

Question

dear folks,
here below is my ad-hoc sql query
SELECT a.ith_po_refno,
a.ith_refno,
a.ith_tdate,
a.ith_vend_code,
a.ith_invno,
pr_whse_name,
pr_whse_code,
cr_acct_name,
ith_bdamt,
(SELECT SUM(b.itd_ibdamt)
FROM itd b,
ith c
WHERE c.ith_refno = b.itd_refno
AND c.ith_po_refno = a.ith_po_refno
AND c.ith_intype = 'R')
totamtr,
(SELECT SUM(d.itd_ibdamt)
FROM itd d,
ith e
WHERE e.ith_refno = d.itd_refno
AND e.ith_po_refno = a.ith_po_refno
AND e.ith_intype = 'Y')
totamty
FROM ith a ,
cr_account,
pr_warehouse
WHERE
a.ith_intype = 'P'
AND a.ith_tdate >= to_date('18/01/2008', 'DD/MM/YYYY')
AND a.ith_tdate <= to_date('18/02/2008', 'DD/MM/YYYY')
AND cr_acct_code = a.ith_vend_code
AND a.ith_ilocn = pr_whse_code
ORDER BY a.ith_add_date ASC
this query runs well in sql developer with 0.0056 secs. the same query takes a longer time in my front end(VB). it takes around 1 minute to process n display it in the list view.
please tel me wat is the problem or how can i optimize the above given query.
thanks all the way
fairozkhan

Answer 1

>how can i optimize the above given query
I'm failing to grasp your logic here. You say you have a query. It runs fast when done via SQL-Developer. The very same query is slow via VB.
Now you want to have suggestions on how to tune the query?
The query is the only constant thing in these two cases. Surely that what is constant is constant. Surely that what is different is resulting in the performance difference?
In other words, it is something that is different between VB and SQL-Developer that results in the very same query to be slower.
It can be that what the VB code does with the results of the query is slow. It can be that the underlaying Oracle driver used by VB has a fetch/array size of 1, and the SQL-Developer a size of a 100. (meaning SQl-Developer fetch data from the Oracle server in bigger and better chunks)
It can be that when you tested it via SQL-Developer, much of the rows to be read where already in the db buffer cache - and when the VB test was done, this data was on disk requiring slow and expensive physical I/O to get it into the buffer cache.
It could be that the SQL-Developer session connected to a dev dabatase (less data) and the VB session to a live database (more data). It could be that the same database are used, but different schemas and thus different environments and scope (which can mean different tables/synonyms/views/FGAC policies/etc)
Or it could sun spots. I heard that these generate a lot of radiation and a burst of these can be "bad" for a computer.
Or it could be that someone step on the network cable to the server throttle the bandwidth when VB was doing its query.
It could be lots of things.. but pointing to the only thing that is common between the two tests, discarding everything else that is different, and saying that which is the same must be causing the difference in performance... I don't get that.
Maybe I just need more coffee?