SQL优化-标量子查询的改写
2016-05-25 15:32
295 查看
原语句：
with d as
(select *
from dc.Dcam_Lbsq d
where d.inptdt between
to_date(substr('20110930', 1, 6) || '01', 'yyyymmdd') and
to_date('20110930', 'yyyymmdd')
and d.lnbltp in ('B', '5', '6')
and d.Begndt <= To_Date('20110930', 'yyyymmdd')
And d.Overdt > To_Date('20110930', 'yyyymmdd'))
select /*+index(c IDX_DCAM_LBSQ_ACCTID) index(d IDX_DCAM_LBSQ_ACCTID)*/
a.lncfno as 账户号,
a.loancn as 合同号,
a.lncfno as 借据编号,
null as 申请号,
b.custno as 客户号,
substr('20110930', 1, 6) as 年月戳,
a.disbdt as 账户开户日期,
a.closdt as 账户关闭日期,
null as 还款频率,
a.initam as 原始贷款金额,
c.onlnbl + nvl(d.onlnbl, 0) as 本期贷款余额,
c.orgibl as 本期应还本金,
nvl(d.orgibl, 0) as 本期应还利息,
c.orgibl - c.onlnbl as 本期实还本金,
nvl(d.orgibl, 0) - nvl(d.onlnbl, 0) as 本期实还利息,
null as 提前还款本金,
case
when a.matudt > '20110930' then
0
else
c.onlnbl
end as 逾期金额,
case
when a.matudt > '20110930' then
0
else
c.onlnbl
end as 本期逾期本金,
nvl(d.onlnbl, 0) as 本期逾期利息,
case
when a.matudt > '20110930' then
0
else
trunc(((to_date(a.matudt, 'yyyymmdd') - to_date('20110930', 'yyyymmdd')) - 1) / 30,
0) + 1
end as 本期逾期期数,
case
when a.matudt > '20110930' then
0
else
to_date('20110930', 'yyyymmdd') - to_date(a.matudt, 'yyyymmdd')
end as 本期最大逾期天数,
case
when a.matudt > '20110930' then
0
else
to_date('20110930', 'yyyymmdd') - to_date(a.matudt, 'yyyymmdd')
end as 逾期天数,
null as 贷款剩余期数,
null as 贷款期数,
(select nvl(sum(i.tranam), 0)
from dc.Dcam_Lscf h
join dc.Dcam_Bspg i
on h.contno = i.contno
where h.lncfno = a.lncfno) as 抵质押物认定价值
from srcapp.lnb_lncf a
join srcapp.knc_acid b
on a.lncfno = b.datavl
--join mon_day j on a.disbdt < '20110930'
join dc.Dcam_Lbsq c
on a.acctid = c.acctid
and c.lnbltp in ('1', '2')
and c.Begndt <= To_Date('20110930', 'yyyymmdd')
And c.Overdt > To_Date('20110930', 'yyyymmdd')
left join d
on a.acctid = d.acctid
where a.retnfs = '1'
and a.disbdt < '20110930'
and a.initam <= 5000000
执行计划：(DCAM_LBSQ有四千多万的数据,KNC_ACID表有一千三百多万的数据,LNB_LNCF 有三十多万的数据 )
13 --------------------------------------------------------------------------------------------------------------------------------------
14 | Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)| Time | Pstart| Pstop |
15 --------------------------------------------------------------------------------------------------------------------------------------
16 | 0 | SELECT STATEMENT | | | | | 3189K(100)| | | |
17 | 1 | SORT AGGREGATE | | 1 | 51 | | | | | |
18 |* 2 | HASH JOIN | | 1 | 51 | | 3197 (2)| 00:00:45 | | |
19 |* 3 | TABLE ACCESS FULL | DCAM_LSCF | 1 | 28 | | 2945 (2)| 00:00:42 | | |
20 | 4 | TABLE ACCESS FULL | DCAM_BSPG | 51086 | 1147K| | 252 (1)| 00:00:04 | | |
21 |* 5 | HASH JOIN | | 2101K| 523M| 79M| 3189K (1)| 12:24:19 | | |
22 |* 6 | HASH JOIN RIGHT OUTER | | 357K| 75M| 4544K| 2244K (1)| 08:43:44 | | |
23 | 7 | PARTITION RANGE ITERATOR | | 77288 | 3622K| | 2197K (1)| 08:32:45 | 1 | 10 |
24 | 8 | VIEW | | 77288 | 3622K| | 2197K (1)| 08:32:45 | | |
25 |* 9 | TABLE ACCESS BY LOCAL INDEX ROWID| DCAM_LBSQ | 77288 | 3622K| | 2197K (1)| 08:32:45 | 1 | 10 |
26 |* 10 | INDEX RANGE SCAN | IDX_DCAM_LBSQ_ACCTID | 1630K| | | 61415 (1)| 00:14:20 | 1 | 10 |
27 |* 11 | HASH JOIN | | 357K| 58M| 51M| 44239 (1)| 00:10:20 | | |
28 |* 12 | TABLE ACCESS FULL | LNB_LNCF | 356K| 47M| | 3481 (2)| 00:00:49 | | |
29 | 13 | TABLE ACCESS FULL | KNC_ACID | 13M| 435M| | 13062 (2)| 00:03:03 | | |
30 | 14 | PARTITION RANGE ITERATOR | | 2097K| 80M| | 937K (1)| 03:38:45 | 1 | 10 |
31 | 15 | TABLE ACCESS BY LOCAL INDEX ROWID | DCAM_LBSQ | 2097K| 80M| | 937K (1)| 03:38:45 | 1 | 10 |
32 |* 16 | INDEX RANGE SCAN | IDX_DCAM_LBSQ_ACCTID | 668K| | | 61383 (1)| 00:14:20 | 1 | 10 |
33 --------------------------------------------------------------------------------------------------------------------------------------
34
35 Predicate Information (identified by operation id):
36 ---------------------------------------------------
37
38 2 - access("H"."CONTNO"="I"."CONTNO")
39 3 - filter(("H"."CONTNO" IS NOT NULL AND "H"."LNCFNO"=:B1))
40 5 - access("A"."ACCTID"="C"."ACCTID")
41 6 - access("A"."ACCTID"="D"."ACCTID")
42 9 - filter(("D"."INPTDT"<=TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "D"."INPTDT">=TO_DATE(' 2011-09-01
43 00:00:00', 'syyyy-mm-dd hh24:mi:ss')))
44 10 - access("D"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "D"."BEGNDT"<=TO_DATE(' 2011-09-30
45 00:00:00', 'syyyy-mm-dd hh24:mi:ss'))
46 filter(("D"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND INTERNAL_FUNCTION("D"."LNBLTP")))
47 11 - access("A"."LNCFNO"="B"."DATAVL")
48 12 - filter(("A"."INITAM"<=5000000 AND "A"."DISBDT"<'20110930' AND "A"."RETNFS"='1'))
49 16 - access("C"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "C"."BEGNDT"<=TO_DATE(' 2011-09-30
50 00:00:00', 'syyyy-mm-dd hh24:mi:ss'))
51 filter(("C"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND INTERNAL_FUNCTION("C"."LNBLTP")))
52 分析说明: 之前还没有使用with as的形式,直接select,根本查不出来数据,使用with as后大概5分钟可以弹出100条数据,但结果集将近30万的数据,预计要5个小时左右才能完全查出结果,速度慢的令人发指.
仔细观察执行计划后,对于大表都走了hash join,但最上面的标量子查询不是如此,于是怀疑性能问题出现在标量子查询部分.
对于标量子查询,如果where谓词条件后面的结果集很少,那么性能没有什么问题,但当where后面结果集很大,比如几十万的数据,那么表量子查询的性能将会很差,这时候考虑将标量子查询改为left join的方式,
让原有的标量子查询中的表 参与表连接.
修改成如下写法后,40秒左右完成查询.
13 ---------------------------------------------------------------------------------------------------------------------------------------
14 | Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)| Time
e0f0
| Pstart| Pstop |
15 ---------------------------------------------------------------------------------------------------------------------------------------
16 | 0 | SELECT STATEMENT | | | | | 3198K(100)| | | |
17 |* 1 | HASH JOIN | | 2101K| 713M| 104M| 3198K (1)| 12:26:26 | | |
18 | 2 | PARTITION RANGE ITERATOR | | 2097K| 80M| | 937K (1)| 03:38:45 | 1 | 10 |
19 | 3 | TABLE ACCESS BY LOCAL INDEX ROWID | DCAM_LBSQ | 2097K| 80M| | 937K (1)| 03:38:45 | 1 | 10 |
20 |* 4 | INDEX RANGE SCAN | IDX_DCAM_LBSQ_ACCTID | 668K| | | 61383 (1)| 00:14:20 | 1 | 10 |
21 |* 5 | HASH JOIN RIGHT OUTER | | 357K| 107M| 8944K| 2252K (1)| 08:45:32 | | |
22 | 6 | VIEW | | 85512 | 7933K| | 3881 (2)| 00:00:55 | | |
23 | 7 | HASH GROUP BY | | 85512 | 4258K| 9M| 3881 (2)| 00:00:55 | | |
24 |* 8 | HASH JOIN | | 85512 | 4258K| | 3198 (2)| 00:00:45 | | |
25 | 9 | TABLE ACCESS FULL | DCAM_BSPG | 51086 | 1147K| | 252 (1)| 00:00:04 | | |
26 |* 10 | TABLE ACCESS FULL | DCAM_LSCF | 86603 | 2368K| | 2944 (2)| 00:00:42 | | |
27 |* 11 | HASH JOIN RIGHT OUTER | | 357K| 75M| 4544K| 2244K (1)| 08:43:44 | | |
28 | 12 | PARTITION RANGE ITERATOR | | 77288 | 3622K| | 2197K (1)| 08:32:45 | 1 | 10 |
29 | 13 | VIEW | | 77288 | 3622K| | 2197K (1)| 08:32:45 | | |
30 |* 14 | TABLE ACCESS BY LOCAL INDEX ROWID| DCAM_LBSQ | 77288 | 3622K| | 2197K (1)| 08:32:45 | 1 | 10 |
31 |* 15 | INDEX RANGE SCAN | IDX_DCAM_LBSQ_ACCTID | 1630K| | | 61415 (1)| 00:14:20 | 1 | 10 |
32 |* 16 | HASH JOIN | | 357K| 58M| 51M| 44239 (1)| 00:10:20 | | |
33 |* 17 | TABLE ACCESS FULL | LNB_LNCF | 356K| 47M| | 3481 (2)| 00:00:49 | | |
34 | 18 | TABLE ACCESS FULL | KNC_ACID | 13M| 435M| | 13062 (2)| 00:03:03 | | |
35 ---------------------------------------------------------------------------------------------------------------------------------------
36
37 Predicate Information (identified by operation id):
38 ---------------------------------------------------
39
40 1 - access("A"."ACCTID"="C"."ACCTID")
41 4 - access("C"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "C"."BEGNDT"<=TO_DATE(' 2011-09-30
42 00:00:00', 'syyyy-mm-dd hh24:mi:ss'))
43 filter(("C"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND INTERNAL_FUNCTION("C"."LNBLTP")))
44 5 - access("A"."LNCFNO"="S"."LNCFNO")
45 8 - access("H"."CONTNO"="I"."CONTNO")
46 10 - filter("H"."CONTNO" IS NOT NULL)
47 11 - access("A"."ACCTID"="D"."ACCTID")
48 14 - filter(("D"."INPTDT"<=TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "D"."INPTDT">=TO_DATE(' 2011-09-01
49 00:00:00', 'syyyy-mm-dd hh24:mi:ss')))
50 15 - access("D"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "D"."BEGNDT"<=TO_DATE(' 2011-09-30
51 00:00:00', 'syyyy-mm-dd hh24:mi:ss'))
52 filter(("D"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND INTERNAL_FUNCTION("D"."LNBLTP")))
53 16 - access("A"."LNCFNO"="B"."DATAVL")
54 17 - filter(("A"."INITAM"<=5000000 AND "A"."DISBDT"<'20110930' AND "A"."RETNFS"='1'))
with d as
(select *
from dc.Dcam_Lbsq d
where d.inptdt between
to_date(substr('20110930', 1, 6) || '01', 'yyyymmdd') and
to_date('20110930', 'yyyymmdd')
and d.lnbltp in ('B', '5', '6')
and d.Begndt <= To_Date('20110930', 'yyyymmdd')
And d.Overdt > To_Date('20110930', 'yyyymmdd'))
select /*+index(c IDX_DCAM_LBSQ_ACCTID) index(d IDX_DCAM_LBSQ_ACCTID)*/
a.lncfno as 账户号,
a.loancn as 合同号,
a.lncfno as 借据编号,
null as 申请号,
b.custno as 客户号,
substr('20110930', 1, 6) as 年月戳,
a.disbdt as 账户开户日期,
a.closdt as 账户关闭日期,
null as 还款频率,
a.initam as 原始贷款金额,
c.onlnbl + nvl(d.onlnbl, 0) as 本期贷款余额,
c.orgibl as 本期应还本金,
nvl(d.orgibl, 0) as 本期应还利息,
c.orgibl - c.onlnbl as 本期实还本金,
nvl(d.orgibl, 0) - nvl(d.onlnbl, 0) as 本期实还利息,
null as 提前还款本金,
case
when a.matudt > '20110930' then
0
else
c.onlnbl
end as 逾期金额,
case
when a.matudt > '20110930' then
0
else
c.onlnbl
end as 本期逾期本金,
nvl(d.onlnbl, 0) as 本期逾期利息,
case
when a.matudt > '20110930' then
0
else
trunc(((to_date(a.matudt, 'yyyymmdd') - to_date('20110930', 'yyyymmdd')) - 1) / 30,
0) + 1
end as 本期逾期期数,
case
when a.matudt > '20110930' then
0
else
to_date('20110930', 'yyyymmdd') - to_date(a.matudt, 'yyyymmdd')
end as 本期最大逾期天数,
case
when a.matudt > '20110930' then
0
else
to_date('20110930', 'yyyymmdd') - to_date(a.matudt, 'yyyymmdd')
end as 逾期天数,
null as 贷款剩余期数,
null as 贷款期数,
(select nvl(sum(i.tranam), 0)
from dc.Dcam_Lscf h
join dc.Dcam_Bspg i
on h.contno = i.contno
where h.lncfno = a.lncfno) as 抵质押物认定价值
from srcapp.lnb_lncf a
join srcapp.knc_acid b
on a.lncfno = b.datavl
--join mon_day j on a.disbdt < '20110930'
join dc.Dcam_Lbsq c
on a.acctid = c.acctid
and c.lnbltp in ('1', '2')
and c.Begndt <= To_Date('20110930', 'yyyymmdd')
And c.Overdt > To_Date('20110930', 'yyyymmdd')
left join d
on a.acctid = d.acctid
where a.retnfs = '1'
and a.disbdt < '20110930'
and a.initam <= 5000000
执行计划：(DCAM_LBSQ有四千多万的数据,KNC_ACID表有一千三百多万的数据,LNB_LNCF 有三十多万的数据 )
13 --------------------------------------------------------------------------------------------------------------------------------------
14 | Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)| Time | Pstart| Pstop |
15 --------------------------------------------------------------------------------------------------------------------------------------
16 | 0 | SELECT STATEMENT | | | | | 3189K(100)| | | |
17 | 1 | SORT AGGREGATE | | 1 | 51 | | | | | |
18 |* 2 | HASH JOIN | | 1 | 51 | | 3197 (2)| 00:00:45 | | |
19 |* 3 | TABLE ACCESS FULL | DCAM_LSCF | 1 | 28 | | 2945 (2)| 00:00:42 | | |
20 | 4 | TABLE ACCESS FULL | DCAM_BSPG | 51086 | 1147K| | 252 (1)| 00:00:04 | | |
21 |* 5 | HASH JOIN | | 2101K| 523M| 79M| 3189K (1)| 12:24:19 | | |
22 |* 6 | HASH JOIN RIGHT OUTER | | 357K| 75M| 4544K| 2244K (1)| 08:43:44 | | |
23 | 7 | PARTITION RANGE ITERATOR | | 77288 | 3622K| | 2197K (1)| 08:32:45 | 1 | 10 |
24 | 8 | VIEW | | 77288 | 3622K| | 2197K (1)| 08:32:45 | | |
25 |* 9 | TABLE ACCESS BY LOCAL INDEX ROWID| DCAM_LBSQ | 77288 | 3622K| | 2197K (1)| 08:32:45 | 1 | 10 |
26 |* 10 | INDEX RANGE SCAN | IDX_DCAM_LBSQ_ACCTID | 1630K| | | 61415 (1)| 00:14:20 | 1 | 10 |
27 |* 11 | HASH JOIN | | 357K| 58M| 51M| 44239 (1)| 00:10:20 | | |
28 |* 12 | TABLE ACCESS FULL | LNB_LNCF | 356K| 47M| | 3481 (2)| 00:00:49 | | |
29 | 13 | TABLE ACCESS FULL | KNC_ACID | 13M| 435M| | 13062 (2)| 00:03:03 | | |
30 | 14 | PARTITION RANGE ITERATOR | | 2097K| 80M| | 937K (1)| 03:38:45 | 1 | 10 |
31 | 15 | TABLE ACCESS BY LOCAL INDEX ROWID | DCAM_LBSQ | 2097K| 80M| | 937K (1)| 03:38:45 | 1 | 10 |
32 |* 16 | INDEX RANGE SCAN | IDX_DCAM_LBSQ_ACCTID | 668K| | | 61383 (1)| 00:14:20 | 1 | 10 |
33 --------------------------------------------------------------------------------------------------------------------------------------
34
35 Predicate Information (identified by operation id):
36 ---------------------------------------------------
37
38 2 - access("H"."CONTNO"="I"."CONTNO")
39 3 - filter(("H"."CONTNO" IS NOT NULL AND "H"."LNCFNO"=:B1))
40 5 - access("A"."ACCTID"="C"."ACCTID")
41 6 - access("A"."ACCTID"="D"."ACCTID")
42 9 - filter(("D"."INPTDT"<=TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "D"."INPTDT">=TO_DATE(' 2011-09-01
43 00:00:00', 'syyyy-mm-dd hh24:mi:ss')))
44 10 - access("D"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "D"."BEGNDT"<=TO_DATE(' 2011-09-30
45 00:00:00', 'syyyy-mm-dd hh24:mi:ss'))
46 filter(("D"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND INTERNAL_FUNCTION("D"."LNBLTP")))
47 11 - access("A"."LNCFNO"="B"."DATAVL")
48 12 - filter(("A"."INITAM"<=5000000 AND "A"."DISBDT"<'20110930' AND "A"."RETNFS"='1'))
49 16 - access("C"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "C"."BEGNDT"<=TO_DATE(' 2011-09-30
50 00:00:00', 'syyyy-mm-dd hh24:mi:ss'))
51 filter(("C"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND INTERNAL_FUNCTION("C"."LNBLTP")))
52 分析说明: 之前还没有使用with as的形式,直接select,根本查不出来数据,使用with as后大概5分钟可以弹出100条数据,但结果集将近30万的数据,预计要5个小时左右才能完全查出结果,速度慢的令人发指.
仔细观察执行计划后,对于大表都走了hash join,但最上面的标量子查询不是如此,于是怀疑性能问题出现在标量子查询部分.
对于标量子查询,如果where谓词条件后面的结果集很少,那么性能没有什么问题,但当where后面结果集很大,比如几十万的数据,那么表量子查询的性能将会很差,这时候考虑将标量子查询改为left join的方式,
让原有的标量子查询中的表 参与表连接.
修改成如下写法后,40秒左右完成查询.
with d as
(select *
from dc.Dcam_Lbsq d
where d.inptdt between
to_date(substr('20110930', 1, 6) || '01', 'yyyymmdd') and
to_date('20110930', 'yyyymmdd')
and d.lnbltp in ('B', '5', '6')
and d.Begndt <= To_Date('20110930', 'yyyymmdd')
And d.Overdt > To_Date('20110930', 'yyyymmdd'))
select /*+index(c IDX_DCAM_LBSQ_ACCTID) index(d IDX_DCAM_LBSQ_ACCTID)*/
a.lncfno as 账户号,
a.loancn as 合同号,
a.lncfno as 借据编号,
null as 申请号,
b.custno as 客户号,
substr('20110930', 1, 6) as 年月戳,
a.disbdt as 账户开户日期,
a.closdt as 账户关闭日期,
null as 还款频率,
a.initam as 原始贷款金额,
c.onlnbl + nvl(d.onlnbl, 0) as 本期贷款余额,
c.orgibl as 本期应还本金,
nvl(d.orgibl, 0) as 本期应还利息,
c.orgibl - c.onlnbl as 本期实还本金,
nvl(d.orgibl, 0) - nvl(d.onlnbl, 0) as 本期实还利息,
null as 提前还款本金,
case
when a.matudt > '20110930' then
0
else
c.onlnbl
end as 逾期金额,
case
when a.matudt > '20110930' then
0
else
c.onlnbl
end as 本期逾期本金,
nvl(d.onlnbl, 0) as 本期逾期利息,
case
when a.matudt > '20110930' then
0
else
trunc(((to_date(a.matudt, 'yyyymmdd') - to_date('20110930', 'yyyymmdd')) - 1) / 30,
0) + 1
end as 本期逾期期数,
case
when a.matudt > '20110930' then
0
else
to_date('20110930', 'yyyymmdd') - to_date(a.matudt, 'yyyymmdd')
end as 本期最大逾期天数,
case
when a.matudt > '20110930' then
0
else
to_date('20110930', 'yyyymmdd') - to_date(a.matudt, 'yyyymmdd')
end as 逾期天数,
null as 贷款剩余期数,
null as 贷款期数,
抵质押物认定价值
from srcapp.lnb_lncf a
join srcapp.knc_acid b
on a.lncfno = b.datavl
--join mon_day j on a.disbdt < '20110930'
join dc.Dcam_Lbsq c
on a.acctid = c.acctid
and c.lnbltp in ('1', '2')
and c.Begndt <= To_Date('20110930', 'yyyymmdd')
And c.Overdt > To_Date('20110930', 'yyyymmdd')
left join d
on a.acctid = d.acctid
left join (select h.lncfno,nvl(sum(i.tranam), 0) as 抵质押物认定价值
from dc.Dcam_Lscf h
join dc.Dcam_Bspg i
on h.contno = i.contno
group by h.lncfno) s
on a.lncfno = s.lncfno
where a.retnfs = '1'
and a.disbdt < '20110930'
and a.initam <= 5000000新的执行计划: 这样看了,这条执行计划是多么的优美...咳咳.13 ---------------------------------------------------------------------------------------------------------------------------------------
14 | Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)| Time
e0f0
| Pstart| Pstop |
15 ---------------------------------------------------------------------------------------------------------------------------------------
16 | 0 | SELECT STATEMENT | | | | | 3198K(100)| | | |
17 |* 1 | HASH JOIN | | 2101K| 713M| 104M| 3198K (1)| 12:26:26 | | |
18 | 2 | PARTITION RANGE ITERATOR | | 2097K| 80M| | 937K (1)| 03:38:45 | 1 | 10 |
19 | 3 | TABLE ACCESS BY LOCAL INDEX ROWID | DCAM_LBSQ | 2097K| 80M| | 937K (1)| 03:38:45 | 1 | 10 |
20 |* 4 | INDEX RANGE SCAN | IDX_DCAM_LBSQ_ACCTID | 668K| | | 61383 (1)| 00:14:20 | 1 | 10 |
21 |* 5 | HASH JOIN RIGHT OUTER | | 357K| 107M| 8944K| 2252K (1)| 08:45:32 | | |
22 | 6 | VIEW | | 85512 | 7933K| | 3881 (2)| 00:00:55 | | |
23 | 7 | HASH GROUP BY | | 85512 | 4258K| 9M| 3881 (2)| 00:00:55 | | |
24 |* 8 | HASH JOIN | | 85512 | 4258K| | 3198 (2)| 00:00:45 | | |
25 | 9 | TABLE ACCESS FULL | DCAM_BSPG | 51086 | 1147K| | 252 (1)| 00:00:04 | | |
26 |* 10 | TABLE ACCESS FULL | DCAM_LSCF | 86603 | 2368K| | 2944 (2)| 00:00:42 | | |
27 |* 11 | HASH JOIN RIGHT OUTER | | 357K| 75M| 4544K| 2244K (1)| 08:43:44 | | |
28 | 12 | PARTITION RANGE ITERATOR | | 77288 | 3622K| | 2197K (1)| 08:32:45 | 1 | 10 |
29 | 13 | VIEW | | 77288 | 3622K| | 2197K (1)| 08:32:45 | | |
30 |* 14 | TABLE ACCESS BY LOCAL INDEX ROWID| DCAM_LBSQ | 77288 | 3622K| | 2197K (1)| 08:32:45 | 1 | 10 |
31 |* 15 | INDEX RANGE SCAN | IDX_DCAM_LBSQ_ACCTID | 1630K| | | 61415 (1)| 00:14:20 | 1 | 10 |
32 |* 16 | HASH JOIN | | 357K| 58M| 51M| 44239 (1)| 00:10:20 | | |
33 |* 17 | TABLE ACCESS FULL | LNB_LNCF | 356K| 47M| | 3481 (2)| 00:00:49 | | |
34 | 18 | TABLE ACCESS FULL | KNC_ACID | 13M| 435M| | 13062 (2)| 00:03:03 | | |
35 ---------------------------------------------------------------------------------------------------------------------------------------
36
37 Predicate Information (identified by operation id):
38 ---------------------------------------------------
39
40 1 - access("A"."ACCTID"="C"."ACCTID")
41 4 - access("C"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "C"."BEGNDT"<=TO_DATE(' 2011-09-30
42 00:00:00', 'syyyy-mm-dd hh24:mi:ss'))
43 filter(("C"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND INTERNAL_FUNCTION("C"."LNBLTP")))
44 5 - access("A"."LNCFNO"="S"."LNCFNO")
45 8 - access("H"."CONTNO"="I"."CONTNO")
46 10 - filter("H"."CONTNO" IS NOT NULL)
47 11 - access("A"."ACCTID"="D"."ACCTID")
48 14 - filter(("D"."INPTDT"<=TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "D"."INPTDT">=TO_DATE(' 2011-09-01
49 00:00:00', 'syyyy-mm-dd hh24:mi:ss')))
50 15 - access("D"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND "D"."BEGNDT"<=TO_DATE(' 2011-09-30
51 00:00:00', 'syyyy-mm-dd hh24:mi:ss'))
52 filter(("D"."OVERDT">TO_DATE(' 2011-09-30 00:00:00', 'syyyy-mm-dd hh24:mi:ss') AND INTERNAL_FUNCTION("D"."LNBLTP")))
53 16 - access("A"."LNCFNO"="B"."DATAVL")
54 17 - filter(("A"."INITAM"<=5000000 AND "A"."DISBDT"<'20110930' AND "A"."RETNFS"='1'))
内容来自用户分享和网络整理，不保证内容的准确性，如有侵权内容，可联系管理员处理 
相关文章推荐
- 解决Mysql下中文乱码全是问号的问题
- Oracle 12c创建用户是出现“ORA-65096: invalid common user or role name”的错误
- 数据库系统工程师 百度脑图
- SQL优化-利用with as的方式去掉nestloop连接
- Dingo + Laravel + JWT + Entrust + memcache 实现API设计
- mysql清屏办法及部分不支持的sql语句
- (error) MISCONF Redis is configured to save RDB snapshots, but is currently not able to persist on d
- mysql etl过程 嵌套3层循环 进行sql优化（欧莱雅视频墙项目）
- 网吧计费系统数据库修复/网吧收银系统数据库恢复
- zabbix 2.2 使用自带的模板监控mysql
- mysql优化
- mysql JDBC URL格式各个参数详解
- Mysql初始化root密码和允许远程访问
- Spark sql处理数据倾斜方法
- 关于oracle数据库访问报12519异常及修改oracle最大连接数的解决方法
- MySQL登录问题
- sql语句的优化
- MySQL 加锁处理分析
- SQL Server执行计划的理解
- 如何借助 NoSQL 提高 JPA 应用性能