I am working to further optimize the following query. I have already had a lot of success optimizing the query up to this point, but now I am looking for help from others for ideas. I appreciate your help.
The following query takes over 200 seconds to complete yet the two individual WITH clause queries (receiving_1 and receiving_2) that make up the "UNION ALL" take less than 4 seconds each.
I was hoping "rownum" would help optimize the "UNION ALL", but it did not. This attempt was based on a note from AskTom --> Files --> SQLTechniques.zip (between 61 and 72)
When you have two queries that run light speed separately But not so together Generally a mixed “CBO/RBO” problem Use of RBO with a feature that kicks in the CBO Rownum can be a temporary fix till all things are CBO
NOTE: One optimization that avoids this issue is embedded replacement parameters or cursor parameters that greatly reduce the rows at key points in the query. However, because I need to run this query through a database link against a read-only database, the following options are not available for this query.
NOTE: In an attempt at providing a good test case with as little code as possible, the below SQL (4,584 chars, 88 line breaks) is a greatly simplified version of the actual SQL (93,683 chars, 2,014 line breaks) I am developing. The original SQL has already gone through several rounds of optimizations to bring the runtime down from over an hour to under 10 seconds. This latest challenge is that the SQL has to be rewritten as a view as opposed to a SQL file with embedded substitution parameters. The substitution parameters were key to the optimizations up to now. I am now refactoring the SQL to work as a view in Oracle 11gR2 to be called via a database link from Oracle 19c. A view has no parameters, obviously. The driving parameters (ledger, OU, period) will be WHERE clause conditions when querying the view.
Big picture: The actual SQL under development is an amalgamation of 6 report queries that share a lot of duplicated code. My work is to convert all six source report queries into an efficient extract for importing into a new database schema for reconciliation purposes.
with
gccs as (
select
gcc.code_combination_id
from apps.gl_code_combinations gcc
where 1=1
AND regexp_like(gcc.segment1,'^[0-9]{4}$') -- only include 4-digit numbers
AND gcc.segment1 BETWEEN to_char('1500') and to_char('1504')
AND regexp_like(gcc.segment3,'^[0-9]{4}$') -- only include 4-digit numbers
AND ( gcc.segment3 BETWEEN to_char('5000') and to_char('5999')
OR gcc.segment3 in ('6000','6002','6003','6004','6005','6006','6007','6008','6009','6010','6011','6012','6013','6014','6017','6018','6019','6020','6021','6022','6023','6024','6025','6026','6027','6028','6029','6030','6031','6032',
'6033','6034','6035','6036','6037','6038','6039','6040','6041','6042','6043','6044','6045','6046','6047','6048','6049','6050','6051','6052','6053','6054','6055','6056','6058','6060','6061','6062','6063','6064',
'6065','6066','6067','6068','6069','6070','6071','6072','6073','6074','6075','6076','6077','6084','6085','6086','6087','6088','6089','6090','6091','6092','6093','6094','6095','6096','6097','6098','6099','6100',
'6101','6102','6103','6104','6105','6106','6120','6121','6130','6131','6140','6150','6151','6155','6158','6162','6990','6991','6992','6993','6994','6995','6996','6997','6998'))
AND regexp_like(gcc.segment6,'^[0-9]{4}$') -- only include 4-digit numbers
AND gcc.segment6 BETWEEN to_char('3000') and to_char('4999')
) -- select count(*) code_combination_id from gccs;
,
base_query as (
SELECT
gl.LEDGER_ID,
hou.organization_id,
gper.PERIOD_NAME,
XAL.AE_HEADER_ID,
XAL.AE_LINE_NUM,
GJB.STATUS BATCH_STATUS,
GJH.JE_CATEGORY,
XAL.ACCOUNTING_CLASS_CODE,
GJH.JE_SOURCE,
GJH.JE_HEADER_ID,
GJL.JE_LINE_NUM
FROM apps.gl_ledgers gl
join apps.hr_operating_units hou
on hou.set_of_books_id = gl.ledger_id
join APPS.GL_JE_LINES GJL
on exists (select null from gccs where gccs.CODE_COMBINATION_ID = GJL.CODE_COMBINATION_ID)
AND (nvl(gjl.stat_amount,0) <> 0 OR (NVL (gjl.accounted_dr, 0) - NVL (gjl.accounted_cr, 0)) <> 0)
join APPS.GL_PERIODS gper
on GJL.PERIOD_NAME = gper.PERIOD_NAME
join APPS.GL_JE_HEADERS GJH
on GJH.JE_HEADER_ID = GJL.JE_HEADER_ID
AND GJH.PERIOD_NAME = gper.PERIOD_NAME
AND GJH.LEDGER_ID = gl.ledger_id
AND GJH.JE_SOURCE = 'Cost Management'
join APPS.GL_JE_BATCHES GJB
on GJB.JE_BATCH_ID = GJH.JE_BATCH_ID
join APPS.GL_IMPORT_REFERENCES GIR
on GIR.JE_HEADER_ID = GJL.JE_HEADER_ID
AND GIR.JE_LINE_NUM = GJL.JE_LINE_NUM
join APPS.XLA_AE_LINES XAL
on XAL.GL_SL_LINK_TABLE = GIR.GL_SL_LINK_TABLE
AND XAL.GL_SL_LINK_ID = GIR.GL_SL_LINK_ID
AND XAL.LEDGER_ID = gl.LEDGER_ID
),
receiving_1 as (
select distinct
bq.LEDGER_ID,
bq.organization_id,
bq.period_name,
bq.AE_HEADER_ID,
bq.AE_LINE_NUM
FROM base_query bq
WHERE bq.BATCH_STATUS = 'P'
AND UPPER(bq.JE_CATEGORY) = UPPER('Receiving')
),
receiving_2 as (
select distinct
bq.LEDGER_ID,
bq.organization_id,
bq.period_name,
bq.AE_HEADER_ID,
bq.AE_LINE_NUM
FROM base_query bq
WHERE bq.ACCOUNTING_CLASS_CODE = 'INTRA'
AND bq.BATCH_STATUS = 'P'
AND UPPER(bq.JE_CATEGORY) = UPPER('Receiving')
),
receiving_all as (
-- the use of "rownum" below sometimes helps to optimize "union all", but does not help in this query
select rownum, x.* from receiving_1 x -- 58 recs in 3.4 secs cost 295 (no UNION)
union all -- 58 recs in 203.1 secs cost 300 (UNION ALL)
select rownum, x.* from receiving_2 x -- 0 recs in 3.8 secs cost 295 (no UNION)
)
select count(*)
from receiving_all
where ledger_id = 2021
and organization_id = 82
and period_name = 'SEP-2020';
AUTOTRACE for receiving_1 part of query"
COUNT(*)
----------
58
Execution Plan
----------------------------------------------------------
-------------------------------------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Pstart| Pstop |
-------------------------------------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 35 | 295 (1)| | |
| 1 | SORT AGGREGATE | | 1 | 35 | | | |
| 2 | VIEW | | 1 | 35 | 295 (1)| | |
| 3 | COUNT | | | | | | |
| 4 | VIEW | | 1 | 35 | 295 (1)| | |
| 5 | HASH UNIQUE | | 1 | 266 | 295 (1)| | |
| 6 | NESTED LOOPS | | 1 | 266 | 294 (1)| | |
| 7 | NESTED LOOPS | | 1 | 266 | 294 (1)| | |
| 8 | NESTED LOOPS | | 1 | 237 | 236 (1)| | |
| 9 | NESTED LOOPS | | 1 | 214 | 178 (1)| | |
| 10 | NESTED LOOPS | | 1 | 207 | 176 (1)| | |
| 11 | NESTED LOOPS | | 1 | 183 | 176 (1)| | |
| 12 | NESTED LOOPS | | 1 | 152 | 173 (1)| | |
| 13 | NESTED LOOPS | | 1 | 148 | 173 (1)| | |
| 14 | NESTED LOOPS | | 1 | 123 | 172 (1)| | |
| 15 | NESTED LOOPS | | 1 | 101 | 169 (1)| | |
| 16 | NESTED LOOPS | | 1 | 97 | 169 (1)| | |
| 17 | NESTED LOOPS | | 1 | 53 | 167 (1)| | |
| 18 | SORT UNIQUE | | 1 | 21 | 22 (0)| | |
| 19 | TABLE ACCESS BY INDEX ROWID| GL_CODE_COMBINATIONS | 1 | 21 | 22 (0)| | |
| 20 | INDEX RANGE SCAN | GL_CODE_COMBINATIONS_N1 | 117 | | 8 (0)| | |
| 21 | TABLE ACCESS BY INDEX ROWID | GL_JE_LINES | 16 | 512 | 144 (0)| | |
| 22 | INDEX RANGE SCAN | GL_JE_LINES_N1 | 319 | | 4 (0)| | |
| 23 | TABLE ACCESS BY INDEX ROWID | GL_JE_HEADERS | 1 | 44 | 2 (0)| | |
| 24 | INDEX UNIQUE SCAN | GL_JE_HEADERS_U1 | 1 | | 1 (0)| | |
| 25 | INDEX UNIQUE SCAN | GL_LEDGERS_U2 | 1 | 4 | 0 (0)| | |
| 26 | TABLE ACCESS BY INDEX ROWID | HR_ORGANIZATION_INFORMATION | 1 | 22 | 3 (0)| | |
| 27 | INDEX RANGE SCAN | HR_ORGANIZATION_INFORMATIO_FK1 | 30 | | 1 (0)| | |
| 28 | INDEX RANGE SCAN | GL_PERIODS_U1 | 1 | 25 | 1 (0)| | |
| 29 | INDEX UNIQUE SCAN | HR_ORGANIZATION_UNITS_PK | 1 | 4 | 0 (0)| | |
| 30 | TABLE ACCESS BY INDEX ROWID | HR_ORGANIZATION_INFORMATION | 1 | 31 | 3 (0)| | |
| 31 | INDEX RANGE SCAN | HR_ORGANIZATION_INFORMATIO_FK2 | 2 | | 1 (0)| | |
| 32 | INDEX UNIQUE SCAN | HR_ALL_ORGANIZATION_UNTS_TL_PK | 1 | 24 | 0 (0)| | |
| 33 | TABLE ACCESS BY INDEX ROWID | GL_JE_BATCHES | 1 | 7 | 2 (0)| | |
| 34 | INDEX UNIQUE SCAN | GL_JE_BATCHES_U1 | 1 | | 1 (0)| | |
| 35 | TABLE ACCESS BY INDEX ROWID | GL_IMPORT_REFERENCES | 105 | 2415 | 58 (0)| | |
| 36 | INDEX RANGE SCAN | GL_IMPORT_REFERENCES_N1 | 105 | | 3 (0)| | |
| 37 | PARTITION LIST ALL | | 1 | | 57 (0)| 1 | 19 |
| 38 | INDEX RANGE SCAN | XLA_AE_LINES_N4 | 1 | | 57 (0)| 1 | 19 |
| 39 | TABLE ACCESS BY LOCAL INDEX ROWID | XLA_AE_LINES | 1 | 29 | 58 (0)| 1 | 1 |
-------------------------------------------------------------------------------------------------------------------------------------
Note
-----
- 'PLAN_TABLE' is old version
Statistics
----------------------------------------------------------
99 recursive calls
0 db block gets
3500320 consistent gets
0 physical reads
116 redo size
346 bytes sent via SQL*Net to client
4388 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
1 sorts (memory)
0 sorts (disk)
1 rows processed
AUTOTRACE for receiving_2 part of query:
COUNT(*)
----------
0
Execution Plan
----------------------------------------------------------
-------------------------------------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Pstart| Pstop |
-------------------------------------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 35 | 295 (1)| | |
| 1 | SORT AGGREGATE | | 1 | 35 | | | |
| 2 | VIEW | | 1 | 35 | 295 (1)| | |
| 3 | COUNT | | | | | | |
| 4 | VIEW | | 1 | 35 | 295 (1)| | |
| 5 | HASH UNIQUE | | 1 | 279 | 295 (1)| | |
| 6 | NESTED LOOPS | | 1 | 279 | 294 (1)| | |
| 7 | NESTED LOOPS | | 1 | 279 | 294 (1)| | |
| 8 | NESTED LOOPS | | 1 | 272 | 292 (1)| | |
| 9 | NESTED LOOPS | | 1 | 230 | 234 (1)| | |
| 10 | NESTED LOOPS | | 1 | 207 | 176 (1)| | |
| 11 | NESTED LOOPS | | 1 | 183 | 176 (1)| | |
| 12 | NESTED LOOPS | | 1 | 152 | 173 (1)| | |
| 13 | NESTED LOOPS | | 1 | 148 | 173 (1)| | |
| 14 | NESTED LOOPS | | 1 | 126 | 170 (1)| | |
| 15 | NESTED LOOPS | | 1 | 122 | 170 (1)| | |
| 16 | NESTED LOOPS | | 1 | 97 | 169 (1)| | |
| 17 | NESTED LOOPS | | 1 | 53 | 167 (1)| | |
| 18 | SORT UNIQUE | | 1 | 21 | 22 (0)| | |
| 19 | TABLE ACCESS BY INDEX ROWID| GL_CODE_COMBINATIONS | 1 | 21 | 22 (0)| | |
| 20 | INDEX RANGE SCAN | GL_CODE_COMBINATIONS_N1 | 117 | | 8 (0)| | |
| 21 | TABLE ACCESS BY INDEX ROWID | GL_JE_LINES | 16 | 512 | 144 (0)| | |
| 22 | INDEX RANGE SCAN | GL_JE_LINES_N1 | 319 | | 4 (0)| | |
| 23 | TABLE ACCESS BY INDEX ROWID | GL_JE_HEADERS | 1 | 44 | 2 (0)| | |
| 24 | INDEX UNIQUE SCAN | GL_JE_HEADERS_U1 | 1 | | 1 (0)| | |
| 25 | INDEX RANGE SCAN | GL_PERIODS_U1 | 1 | 25 | 1 (0)| | |
| 26 | INDEX UNIQUE SCAN | GL_LEDGERS_U2 | 1 | 4 | 0 (0)| | |
| 27 | TABLE ACCESS BY INDEX ROWID | HR_ORGANIZATION_INFORMATION | 1 | 22 | 3 (0)| | |
| 28 | INDEX RANGE SCAN | HR_ORGANIZATION_INFORMATIO_FK1 | 30 | | 1 (0)| | |
| 29 | INDEX UNIQUE SCAN | HR_ORGANIZATION_UNITS_PK | 1 | 4 | 0 (0)| | |
| 30 | TABLE ACCESS BY INDEX ROWID | HR_ORGANIZATION_INFORMATION | 1 | 31 | 3 (0)| | |
| 31 | INDEX RANGE SCAN | HR_ORGANIZATION_INFORMATIO_FK2 | 2 | | 1 (0)| | |
| 32 | INDEX UNIQUE SCAN | HR_ALL_ORGANIZATION_UNTS_TL_PK | 1 | 24 | 0 (0)| | |
| 33 | TABLE ACCESS BY INDEX ROWID | GL_IMPORT_REFERENCES | 105 | 2415 | 58 (0)| | |
| 34 | INDEX RANGE SCAN | GL_IMPORT_REFERENCES_N1 | 105 | | 3 (0)| | |
| 35 | PARTITION LIST ALL | | 1 | 42 | 58 (0)| 1 | 19 |
| 36 | TABLE ACCESS BY LOCAL INDEX ROWID | XLA_AE_LINES | 1 | 42 | 58 (0)| 1 | 19 |
| 37 | INDEX RANGE SCAN | XLA_AE_LINES_N4 | 1 | | 57 (0)| 1 | 19 |
| 38 | INDEX UNIQUE SCAN | GL_JE_BATCHES_U1 | 1 | | 1 (0)| | |
| 39 | TABLE ACCESS BY INDEX ROWID | GL_JE_BATCHES | 1 | 7 | 2 (0)| | |
-------------------------------------------------------------------------------------------------------------------------------------
Note
-----
- 'PLAN_TABLE' is old version
Statistics
----------------------------------------------------------
99 recursive calls
0 db block gets
3451912 consistent gets
0 physical reads
0 redo size
345 bytes sent via SQL*Net to client
4385 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
1 sorts (memory)
0 sorts (disk)
1 rows processed
AUTOTRACE for UNION ALL of full query that includes both request_1 and request_2
COUNT(*)
----------
58
Execution Plan
----------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Pstart| Pstop |
----------------------------------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 35 | 300 (1)| | |
| 1 | TEMP TABLE TRANSFORMATION | | | | | | |
| 2 | LOAD AS SELECT | SYS_TEMP_0FD9D6767_FC3F7B00 | | | | | |
| 3 | NESTED LOOPS | | 1 | 279 | 294 (1)| | |
| 4 | NESTED LOOPS | | 1 | 279 | 294 (1)| | |
| 5 | NESTED LOOPS | | 1 | 237 | 236 (1)| | |
| 6 | NESTED LOOPS | | 1 | 214 | 178 (1)| | |
| 7 | NESTED LOOPS | | 1 | 207 | 176 (1)| | |
| 8 | NESTED LOOPS | | 1 | 183 | 176 (1)| | |
| 9 | NESTED LOOPS | | 1 | 152 | 173 (1)| | |
| 10 | NESTED LOOPS | | 1 | 148 | 173 (1)| | |
| 11 | NESTED LOOPS | | 1 | 126 | 170 (1)| | |
| 12 | NESTED LOOPS | | 1 | 122 | 170 (1)| | |
| 13 | NESTED LOOPS | | 1 | 78 | 168 (1)| | |
| 14 | NESTED LOOPS | | 1 | 53 | 167 (1)| | |
| 15 | SORT UNIQUE | | 1 | 21 | 22 (0)| | |
| 16 | TABLE ACCESS BY INDEX ROWID| GL_CODE_COMBINATIONS | 1 | 21 | 22 (0)| | |
| 17 | INDEX RANGE SCAN | GL_CODE_COMBINATIONS_N1 | 117 | | 8 (0)| | |
| 18 | TABLE ACCESS BY INDEX ROWID | GL_JE_LINES | 16 | 512 | 144 (0)| | |
| 19 | INDEX RANGE SCAN | GL_JE_LINES_N1 | 319 | | 4 (0)| | |
| 20 | INDEX RANGE SCAN | GL_PERIODS_U1 | 1 | 25 | 1 (0)| | |
| 21 | TABLE ACCESS BY INDEX ROWID | GL_JE_HEADERS | 1 | 44 | 2 (0)| | |
| 22 | INDEX UNIQUE SCAN | GL_JE_HEADERS_U1 | 1 | | 1 (0)| | |
| 23 | INDEX UNIQUE SCAN | GL_LEDGERS_U2 | 1 | 4 | 0 (0)| | |
| 24 | TABLE ACCESS BY INDEX ROWID | HR_ORGANIZATION_INFORMATION | 1 | 22 | 3 (0)| | |
| 25 | INDEX RANGE SCAN | HR_ORGANIZATION_INFORMATIO_FK1 | 30 | | 1 (0)| | |
| 26 | INDEX UNIQUE SCAN | HR_ORGANIZATION_UNITS_PK | 1 | 4 | 0 (0)| | |
| 27 | TABLE ACCESS BY INDEX ROWID | HR_ORGANIZATION_INFORMATION | 1 | 31 | 3 (0)| | |
| 28 | INDEX RANGE SCAN | HR_ORGANIZATION_INFORMATIO_FK2 | 2 | | 1 (0)| | |
| 29 | INDEX UNIQUE SCAN | HR_ALL_ORGANIZATION_UNTS_TL_PK | 1 | 24 | 0 (0)| | |
| 30 | TABLE ACCESS BY INDEX ROWID | GL_JE_BATCHES | 1 | 7 | 2 (0)| | |
| 31 | INDEX UNIQUE SCAN | GL_JE_BATCHES_U1 | 1 | | 1 (0)| | |
| 32 | TABLE ACCESS BY INDEX ROWID | GL_IMPORT_REFERENCES | 105 | 2415 | 58 (0)| | |
| 33 | INDEX RANGE SCAN | GL_IMPORT_REFERENCES_N1 | 105 | | 3 (0)| | |
| 34 | PARTITION LIST ALL | | 1 | | 57 (0)| 1 | 19 |
| 35 | INDEX RANGE SCAN | XLA_AE_LINES_N4 | 1 | | 57 (0)| 1 | 19 |
| 36 | TABLE ACCESS BY LOCAL INDEX ROWID | XLA_AE_LINES | 1 | 42 | 58 (0)| 1 | 1 |
| 37 | SORT AGGREGATE | | 1 | 35 | | | |
| 38 | VIEW | | 2 | 70 | 6 (34)| | |
| 39 | UNION-ALL | | | | | | |
| 40 | COUNT | | | | | | |
| 41 | VIEW | | 1 | 35 | 3 (34)| | |
| 42 | HASH UNIQUE | | 1 | 44 | 3 (34)| | |
| 43 | VIEW | | 1 | 44 | 2 (0)| | |
| 44 | TABLE ACCESS FULL | SYS_TEMP_0FD9D6767_FC3F7B00 | 1 | 74 | 2 (0)| | |
| 45 | COUNT | | | | | | |
| 46 | VIEW | | 1 | 35 | 3 (34)| | |
| 47 | HASH UNIQUE | | 1 | 61 | 3 (34)| | |
| 48 | VIEW | | 1 | 61 | 2 (0)| | |
| 49 | TABLE ACCESS FULL | SYS_TEMP_0FD9D6767_FC3F7B00 | 1 | 74 | 2 (0)| | |
----------------------------------------------------------------------------------------------------------------------------------
Note
-----
- 'PLAN_TABLE' is old version
Statistics
----------------------------------------------------------
766 recursive calls
83076 db block gets
270487689 consistent gets
81660 physical reads
676 redo size
346 bytes sent via SQL*Net to client
4582 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
1 sorts (memory)
0 sorts (disk)
1 rows processed
As I don't know the underlying tables, their structure and their data, I can only guess. Having said that, I'd try the following things (not knowing if any of the steps are helpful or not!):
remove the rownum from receiving_all
add the where clause from the main query to each of the subqueries in receiving_all. The optimizer might be able to infer this, but I am not sure.
change union all to union in receiving_all and remove the distinct from receiving_1 and receiving_2
receiving_2 seems to be included in receiving_2?
If 4. is not the case, I would combine both into a single query with WHERE UPPER(bq.JE_CATEGORY) = UPPER('Receiving') AND ( (<conditions for 1>) OR (<conditions for 2>) )
I would try if UPPER(bq.JE_CATEGORY) can be avoided. Functions prevent index access.
In case you can create tables, I'd unroll the values of gccs into two tables, one for segment_1, and another table for segment_2.
If 7. is not an option, I'd add the hint select /*+ materialize */ gcc.code_combination_id to gcc.
I'm very curious to see which suggestion has any positive impact...