Stream Line SQL Query

Question

Hello all,
I have a table that contains more than 100 000 records, structure seen below. I am trying to perform simple queries on this table and it is taking FOREVER for the query to return (in excess of 10 minutes). The only field that contains unique data is the 'key', which I have made the primary key.
Table: WORK
Name Null? Type
ASSIGNMENT_NO NOT NULL NUMBER(2)
START_DATE NOT NULL DATE
END_DATE DATE
ID_NO NOT NULL NUMBER(9)
CODE NOT NULL VARCHAR2(4)
REF_KEY NOT NULL NUMBER(9)
STATUS NOT NULL VARCHAR2(1)
ID_NAME NOT NULL VARCHAR2(30)
LAST_MODIFIED_DATE NOT NULL DATE
KEY NOT NULL NUMBER(9)
As I need to pass into the query varying data I Use a PL/SQL function to call the SQL query.
CREATE OR REPLACE FUNCTION THE_SAMPLE (THE_REF_KEY, THE_ID_NO, PUT_A_DATE_HERE)
RETURN VARCHAR2 IS
CURSOR C IS
SELECT w1.code
FROM work w1
WHERE w1.ref_key = THE_REF_KEY
AND w1.ID_NO = THE_ID_NO
AND w1.status <> 'I'
AND w1.start_date <= PUT_A_DATE_HERE
AND (w1.end_date IS NULL or w1.end_date >= PUT_A_DATE_HERE)
AND w1.start_date = (SELECT MAX(W2.start_date)
FROM work w2
WHERE w2.REF_KEY = W1.REF_KEY
AND w2.ID_NO = W1.ID_NO
AND w2.status <> 'I'
AND w2.start_date <= PUT_A_DATE_HERE
AND (w2.end_date IS NULL or w2.end_date >= PUT_A_DATE_HERE));
out_rec c%ROWTYPE;
begin
begin
open C;
fetch c into out_rec;
CLOSE C;
EXCEPTION
WHEN no_data_found THEN
out_rec.code := NULL;
END;
return out_rec.code;
end THE_SAMPLE;
I am looking for a quick way to accomplish the same task. The main reason is that I am feeding this SQL statement about 20 000 different ref_keys each with a particular ID_NO. Since it seems to take a long time just to process one element I did want crunch data for the next month.
Any help, idea, comments would be greatly welcomed.
Any ideas on how to stream line this?

Answer 1

Hummm... I am still working on it, but this is what I have come up with so far:
One way is to do this use this
CURSOR C IS
SELECT w1.code
FROM work w1, (SELECT MAX(w2.start_date) as MAX_DATE
FROM work w2
WHERE w2.ref_key = THE_REF_KEY
AND w2.ID_NO = THE_ID_NO
and w2.status <> 'I'
AND w2.start_date <= PUT_A_DATE_HERE
AND (w2.end_date IS NULL or w2.end_date >= PUT_A_DATE_HERE)) GET_MAX
WHERE w1.ref_key = THE_REF_KEY
AND w1.ID_NO = THE_ID_NO
and w1.status <> 'I'
AND w1.start_date <= PUT_A_DATE_HERE
AND (w1.end_date IS NULL or w1.end_date >= PUT_A_DATE_HERE)
AND w1.start_date = GET_MAX.MAX_DATE;
This way does seem to work a bit faster. I will have to see as the hours pass by.
I am still looking for ways that will process this faster.
null