DISTINCT operator performance issue

Question

Hi Guyz,
I am facing a performance issue in a query which contains DISTINCT function. Following is my query:
SELECT     /*+ ORDERED USE_NL_WITH_INDEX(c DIMENSION_KEY_PK) */
                    DISTINCT f.*,c.client_ids
FROM FACT_TAB f, DIM_TAB c
WHERE f.client = c.dimension_key
FACT_TAB = Fact table with a bitmap index on client column (10,000,000 records).
DIM_TAB = Dimension table with dimension_key as primary key (100,000 records).
when i select only fact table columns in the above query, the query executes within a second. But when i execute the above query it takes more than 15 minutes to execute.
How can i improve the above query. Any suggestions or tips would be helpful.
Thanks in advance.

Answer 1

Hi myers,
you are absolutely right, there is no purpose of using DISTINCT, because i have found there are no duplicates in fact table, neither in dimension.
BUT there is another problem after this, i am joining these two tables with another table (INLINE VIEW), which gives me duplicate data, so i need DISTINCT operator for that purpose. Time dimension is also used in this query now. Here is my new Query:
SELECT /*+ ORDERED USE_NL_WITH_INDEX(c DIMENSION_KEY_PK) */
DISTINCT f.*,c.client_ids
FROM FACT_TAB f, DIM_TAB c, DIM_TIME_TAB t, (select id,start_date,end_date from tab3) tab
WHERE f.client = c.dimension_key
AND f.time = t.dimension_key
AND f.tabid = tab.id
AND t.day_start_date >= tab.start_date
AND (t.day_start_date <= tab.end_date OR tab.end_date IS NULL)
Thanks