Query not working fine

Question

Hi
db version is 9.2.0.8.0
in dev the query is working fine.But in qa the query is not working fine(hanging).I had checked the explain plans.
Both are different.This is due to difference in volume of data.
I am pasting the explain plans of both the queries in qa and dev
qa explain plan
| Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)|
| 0 | SELECT STATEMENT | | 606M| 118G| | 107M (1)|
| 1 | SORT GROUP BY | | 606M| 118G| 524G| 107M (1)|
| 2 | HASH JOIN | | 1869M| 363G| | 50748 (1)|
| 3 | TABLE ACCESS FULL | DIM_GEOGRAPHY_HIER | 9066 | 318K| | 331 (1)|
| 4 | HASH JOIN | | 2474K| 408M| 2232K| 50400 (1)|
| 5 | TABLE ACCESS FULL | DIM_GEOGRAPHY_HIER | 54396 | 1593K| | 330 (1)|
| 6 | HASH JOIN | | 2474K| 337M| | 31791 (1)|
| 7 | TABLE ACCESS FULL | DIM_PRODUCT_HIER | 206 | 3296 | | 25 (4)|
| 8 | HASH JOIN | | 3162K| 382M| | 31745 (1)|
| 9 | TABLE ACCESS FULL | DIM_PRODUCT_HIER | 3087 | 27783 | | 25 (4)|
| 10 | HASH JOIN | | 3162K| 355M| 2952K| 31699 (1)|
| 11 | MERGE JOIN CARTESIAN | | 62930 | 2212K| | 14627 (1)|
| 12 | INDEX FAST FULL SCAN | SYS_C0013071 | 2923 | | | 4 (25)|
| 13 | BUFFER SORT | | 22 | 792 | | 14623 (1)|
| 14 | TABLE ACCESS BY INDEX ROWID| CURCY_CONVERT_RATE | 22 | 792 | | 6 (17)|
| 15 | INDEX RANGE SCAN | XPKCURRATE | 22 | | | 5 (20)|
| 16 | TABLE ACCESS FULL | SHIPPABLE_BACKLOG_FACT | 321K| 25M| | 15494 (1)|
dev explain plan
| Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)|
| 0 | SELECT STATEMENT | | 17M| 2978M| | 927K (1)|
| 1 | SORT GROUP BY | | 17M| 2978M| 6504M| 927K (1)|
|* 2 | HASH JOIN | | 17M| 2978M| | 2726 (1)|
|* 3 | TABLE ACCESS FULL | DIM_GEOGRAPHY_HIER | 4992 | 180K| | 416 (1)|
|* 4 | HASH JOIN | | 41051 | 5852K| 2872K| 2309 (1)|
| 5 | TABLE ACCESS FULL | DIM_GEOGRAPHY_HIER | 69888 | 2047K| | 415 (1)|
|* 6 | HASH JOIN | | 41051 | 4650K| | 1504 (1)|
|* 7 | TABLE ACCESS FULL | DIM_PRODUCT_HIER | 206 | 3296 | | 37 (3)|
|* 8 | HASH JOIN | | 52460 | 5123K| | 1467 (1)|
| 9 | TABLE ACCESS FULL | DIM_PRODUCT_HIER | 3087 | 27783 | | 37 (3)|
|* 10 | HASH JOIN | | 52460 | 4661K| | 1430 (1)|
|* 11 | TABLE ACCESS FULL | SHIPPABLE_BACKLOG_FACT | 7718 | 414K| | 513 (1)|
| 12 | MERGE JOIN CARTESIAN | | 48941 | 1720K| | 915 (0)|
| 13 | INDEX FULL SCAN | SYS_C0060330 | 1827 | | | 7 (15)|
| 14 | BUFFER SORT | | 27 | 972 | | 913 (0)|
| 15 | TABLE ACCESS BY INDEX ROWID| CURCY_CONVERT_RATE | 27 | 972 | | 2 (50)|
|* 16 | INDEX RANGE SCAN | XPKCURRATE | 27 | | | 4 (25)|
Here is the query
SELECT 'BACKLOG' AS SUBJECT_AREA, 'CURRENT' AS TIME_RANGE
,GEO_ROLLUP.GEOGRAPHY_HIER_KEY
,PRODUCT_ROLLUP.PRODUCT_HIER_KEY
, backlog_FACT.DISTRIBUTION_CH_ID
, backlog_FACT.SALES_DISTRICT_ID
, SUM(NVL(backlog_FACT.GC_SCHEDULED_AMT,0) - NVL(backlog_FACT.GC_biilled_AMT,0)) AS billing_backlog_AMT
, SUM(NVL(backlog_FACT.GC_SCHEDULED_AMT,0)- NVL(backlog_FACT.GC_DELIVERED_AMT,0)) AS delivery_backlog_AMT
, SUM(NVL(backlog_FACT.GC_SCHEDULED_AMT,0) - NVL(backlog_FACT.GC_REV_REC_AMT,0)) AS rev_rec_backlog_AMT
, CURCY.VALID_FROM
FROM fact.shippable_backlog_FACT backlog_FACT
, DIM.DIM_PRODUCT_HIER PRODUCT
, DIM.DIM_GEOGRAPHY_HIER GEO
, DIM.DIM_DATE DT
, DIM.DIM_PRODUCT_HIER PRODUCT_ROLLUP
, DIM.DIM_GEOGRAPHY_HIER GEO_ROLLUP
, DIM.CURCY_CONVERT_RATE CURCY
WHERE
PRODUCT.PRODUCT_HIER_KEY = backlog_FACT.GEF_PRODUCT_HIER_KEY
AND GEO.GEOGRAPHY_HIER_KEY = backlog_FACT.GEF_SHIP_TO_GEO_HIER_KEY
AND ( backlog_FACT.DELIVERY_COMPLETE_DATE IS NULL
     OR backlog_FACT.BILLING_COMPLETE_DATE IS NULL
     OR backlog_FACT.REV_REC_COMPLETE_DATE IS NULL
AND PRODUCT.PRODUCT_TYPE = PRODUCT_ROLLUP.PRODUCT_TYPE
AND PRODUCT_ROLLUP.HIER_LEVEL_NUM = 4
AND PRODUCT_ROLLUP.CURRENT_RECORD_IND = 'Y'
AND GEO_ROLLUP.SUB_POLE = GEO.SUB_POLE
AND GEO_ROLLUP.HIER_LEVEL_NUM = 3
AND GEO_ROLLUP.CURRENT_RECORD_IND = 'Y'
AND CURCY.VALID_FROM <= BACKLOG_FACT.ORDER_LINE_CREATE_DATE
AND CURCY.VALID_TO > BACKLOG_FACT.ORDER_LINE_CREATE_DATE
AND CURCY.EXCHANGE_TYPE_ID = 'M'
AND CURCY.TO_CURRENCY_ID = 'USD'
AND CURCY.FROM_CURRENCY_ID = BACKLOG_FACT.DOCUMENT_CURRENCY_ID
GROUP BY
PRODUCT_ROLLUP.PRODUCT_HIER_KEY,
GEO_ROLLUP.GEOGRAPHY_HIER_KEY,
backlog_FACT.DISTRIBUTION_CH_ID,
backlog_FACT.SALES_DISTRICT_ID,
CURCY.VALID_FROM;
I analyzed all the tables involved in it.
can anybody give me suggestion in crating indexes or anything else?
Thanks
Veer

Answer 1

Veer,
Modify your post and enclose your code and output between \ tags for formatting
\Your code or output goes here
\Now, are table stats are upto date on all the table on QA machine? If not, analyze your table and indexes and re-run your query
Regards
Edited by: OrionNet on May 7, 2009 3:35 PM