[20181225]12CR2 SQL Plan Directives.txt

[20181225]12CR2 SQL Plan Directives.txt


--//12C引入SQL PLAN Directives.12cR1版本会造成大量的动态取样,影响性能.许多人把OPTIMIZER_ADAPTIVE_FEATURES设置为false.

--//这也是为什么我不主张将XX.1版本使用在生产系统.12CR2做了一些改进,废除了OPTIMIZER_ADAPTIVE_FEATURES参数.使用2个新的

--//参数OPTIMIZER_ADAPTIVE_PLANS,OPTIMIZER_ADAPTIVE_STATISTICS,缺省前者true,后者为false.

--//通过测试说明问题.


1.环境:

SCOTT@test01p> @ ver1

PORT_STRING          VERSION    BANNER                                                                               CON_ID

-------------------- ---------- -------------------------------------------------------------------------------- ----------

IBMPC/WIN_NT64-9.1.0 12.2.0.1.0 Oracle Database 12c Enterprise Edition Release 12.2.0.1.0 - 64bit Production              0


SCOTT@test01p> show parameter OPTIMIZER_ADAPTIVE

NAME                              TYPE    VALUE

--------------------------------- ------- ------

optimizer_adaptive_plans          boolean TRUE

optimizer_adaptive_reporting_only boolean FALSE

optimizer_adaptive_statistics     boolean FALSE


--//注:没有OPTIMIZER_ADAPTIVE_FEATURES参数,optimizer_adaptive_plans=true,optimizer_adaptive_statistics=false.


2.建立测试环境:

CREATE TABLE t

AS

       SELECT ROWNUM id

             ,LPAD ('x', 20, 'x') name

             ,MOD (ROWNUM, 3) flag1

             ,MOD (ROWNUM, 3) flag2

             ,MOD (ROWNUM, 3) flag3

         FROM DUAL

   CONNECT BY LEVEL <= 1e5;


--//说明:flags1,flags2,flags3分别存在3个取值,按照道理存在27种选择.因为存在相关性,仅仅存在3种选择.


3.测试:

SCOTT@test01p> alter session set statistics_level=all;

Session altered.


SCOTT@test01p> select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1;

COUNT(DISTINCTNAME)

-------------------

                  1


SCOTT@test01p> @ dpc '' ''

PLAN_TABLE_OUTPUT

-------------------------------------

SQL_ID  872fdta99gdk8, child number 0

-------------------------------------

select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1

Plan hash value: 2359337548

---------------------------------------------------------------------------------------------------------------------------------------------------------------

| Id  | Operation            | Name     | Starts | E-Rows |E-Bytes| Cost (%CPU)| E-Time   | A-Rows |   A-Time   | Buffers | Reads  |  OMem |  1Mem | Used-Mem |

---------------------------------------------------------------------------------------------------------------------------------------------------------------

|   0 | SELECT STATEMENT     |          |      1 |        |       |   155 (100)|          |      1 |00:00:00.06 |     556 |    540 |       |       |          |

|   1 |  SORT AGGREGATE      |          |      1 |      1 |    12 |            |          |      1 |00:00:00.06 |     556 |    540 |       |       |          |

|   2 |   VIEW               | VW_DAG_0 |      1 |      1 |    12 |   155   (2)| 00:00:01 |      1 |00:00:00.06 |     556 |    540 |       |       |          |

|   3 |    HASH GROUP BY     |          |      1 |      1 |    30 |   155   (2)| 00:00:01 |      1 |00:00:00.06 |     556 |    540 |  1345K|  1345K|  504K (0)|

|*  4 |     TABLE ACCESS FULL| T        |      1 |   3704 |   108K|   154   (1)| 00:00:01 |  33334 |00:00:00.06 |     556 |    540 |       |       |          |

---------------------------------------------------------------------------------------------------------------------------------------------------------------

Query Block Name / Object Alias (identified by operation id):

-------------------------------------------------------------

   1 - SEL$C33C846D

   2 - SEL$5771D262 / VW_DAG_0@SEL$C33C846D

   3 - SEL$5771D262

   4 - SEL$5771D262 / T@SEL$1

Predicate Information (identified by operation id):

---------------------------------------------------

   4 - filter(("FLAG1"=1 AND "FLAG2"=1 AND "FLAG3"=1))


--//注意看id=4, E-Rows=3704,估算按照100000/27 = 3703.7,而A-Rows=33334(10000/3 = 3333.3),存在很大偏差.


SCOTT@test01p> select sql_id,child_number,is_reoptimizable  from v$sql  where sql_id ='872fdta99gdk8';

SQL_ID        CHILD_NUMBER I

------------- ------------ -

872fdta99gdk8            0 Y

--//is_reoptimizable='Y'


SCOTT@test01p> exec dbms_spd.flush_sql_plan_directive;

PL/SQL procedure successfully completed.


set numw 20

column NOTES format a50


SELECT directive_id

      ,TYPE

      ,enabled

      ,state

      ,notes

      ,reason

  FROM dba_sql_plan_directives

 WHERE directive_id IN (SELECT directive_id

                          FROM dba_sql_plan_dir_objects

                         WHERE owner = USER AND object_name = 'T');


        DIRECTIVE_ID TYPE                 ENA STATE                NOTES                                              REASON

-------------------- -------------------- --- -------------------- -------------------------------------------------- ------------------------------------

17342821566768621333 DYNAMIC_SAMPLING     YES USABLE               NEW

                                                                   undant>NO{EC(SCOTT.T)[FLAG1,

                                                                    FLAG2, FLAG3]}

--//指导建议字段flag1,flag2,flag3联合查询时存在偏差,建议动态取样.

--//补充说明:{EC(SCOTT.T)[FLAG1,FLAG2, FLAG3]}

--//这里的E和C,以及可能出现其他的字符,解释如下:

E – equality_predicates_only

C – simple_column_predicates_only

J – index_access_by_join_predicates

F – filter_on_joining_object


--//再次执行:


SCOTT@test01p> select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1;

COUNT(DISTINCTNAME)

-------------------

                  1


SCOTT@test01p> @ dpc '' ''

PLAN_TABLE_OUTPUT

-------------------------------------

SQL_ID  872fdta99gdk8, child number 1

-------------------------------------

select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1

Plan hash value: 2359337548

------------------------------------------------------------------------------------------------------------------------------------------------------

| Id  | Operation            | Name     | Starts | E-Rows |E-Bytes| Cost (%CPU)| E-Time   | A-Rows |   A-Time   | Buffers |  OMem |  1Mem | Used-Mem |

------------------------------------------------------------------------------------------------------------------------------------------------------

|   0 | SELECT STATEMENT     |          |      1 |        |       |   156 (100)|          |      1 |00:00:00.01 |     556 |       |       |          |

|   1 |  SORT AGGREGATE      |          |      1 |      1 |    12 |            |          |      1 |00:00:00.01 |     556 |       |       |          |

|   2 |   VIEW               | VW_DAG_0 |      1 |      1 |    12 |   156   (2)| 00:00:01 |      1 |00:00:00.01 |     556 |       |       |          |

|   3 |    HASH GROUP BY     |          |      1 |      1 |    30 |   156   (2)| 00:00:01 |      1 |00:00:00.01 |     556 |  1345K|  1345K|  505K (0)|

|*  4 |     TABLE ACCESS FULL| T        |      1 |  33334 |   976K|   154   (1)| 00:00:01 |  33334 |00:00:00.01 |     556 |       |       |          |

------------------------------------------------------------------------------------------------------------------------------------------------------

Query Block Name / Object Alias (identified by operation id):

-------------------------------------------------------------

   1 - SEL$C33C846D

   2 - SEL$5771D262 / VW_DAG_0@SEL$C33C846D

   3 - SEL$5771D262

   4 - SEL$5771D262 / T@SEL$1

Predicate Information (identified by operation id):

---------------------------------------------------

   4 - filter(("FLAG1"=1 AND "FLAG2"=1 AND "FLAG3"=1))

Note

-----

   - statistics feedback used for this statement

--//注意note,指示statistics feedback used for this statement.


SCOTT@test01p> select sql_id,child_number,is_reoptimizable  from v$sql  where sql_id ='872fdta99gdk8';

SQL_ID                CHILD_NUMBER I

------------- -------------------- -

872fdta99gdk8                    0 Y

872fdta99gdk8                    1 N


SCOTT@test01p> @ share 872fdta99gdk8

SQL_TEXT                       = select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1

SQL_ID                         = 872fdta99gdk8

ADDRESS                        = 000007FF1393F830

CHILD_ADDRESS                  = 000007FF13D9C198

CHILD_NUMBER                   = 0

USE_FEEDBACK_STATS             = Y

REASON                         = 048Auto Reoptimization Mismatch(1)3x43241945600

--------------------------------------------------

SQL_TEXT                       = select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1

SQL_ID                         = 872fdta99gdk8

ADDRESS                        = 000007FF1393F830

CHILD_ADDRESS                  = 000007FF115A7E58

CHILD_NUMBER                   = 1

REASON                         =

--------------------------------------------------

PL/SQL procedure successfully completed.


SELECT directive_id

      ,TYPE

      ,enabled

      ,state

      ,notes

      ,reason

  FROM dba_sql_plan_directives

 WHERE directive_id IN (SELECT directive_id

                          FROM dba_sql_plan_dir_objects

                         WHERE owner = USER AND object_name = 'T');


        DIRECTIVE_ID TYPE                 ENA STATE                NOTES                                              REASON

-------------------- -------------------- --- -------------------- -------------------------------------------------- ------------------------------------

17342821566768621333 DYNAMIC_SAMPLING     YES USABLE               NEW

                                                                   undant>NO{EC(SCOTT.T)[FLAG1,

                                                                    FLAG2, FLAG3]}


4.继续测试:

--//设置OPTIMIZER_ADAPTIVE_STATISTICS=true看看.

SCOTT@test01p> alter session set OPTIMIZER_ADAPTIVE_STATISTICS=true ;

Session altered.


SCOTT@test01p> select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1;

COUNT(DISTINCTNAME)

-------------------

                  1


SCOTT@test01p> @ dpc '' ''

PLAN_TABLE_OUTPUT

-------------------------------------

SQL_ID  872fdta99gdk8, child number 2

-------------------------------------

select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1

Plan hash value: 2359337548

------------------------------------------------------------------------------------------------------------------------------------------------------

| Id  | Operation            | Name     | Starts | E-Rows |E-Bytes| Cost (%CPU)| E-Time   | A-Rows |   A-Time   | Buffers |  OMem |  1Mem | Used-Mem |

------------------------------------------------------------------------------------------------------------------------------------------------------

|   0 | SELECT STATEMENT     |          |      1 |        |       |   157 (100)|          |      1 |00:00:00.01 |     556 |       |       |          |

|   1 |  SORT AGGREGATE      |          |      1 |      1 |    12 |            |          |      1 |00:00:00.01 |     556 |       |       |          |

|   2 |   VIEW               | VW_DAG_0 |      1 |      1 |    12 |   157   (3)| 00:00:01 |      1 |00:00:00.01 |     556 |       |       |          |

|   3 |    HASH GROUP BY     |          |      1 |      1 |    30 |   157   (3)| 00:00:01 |      1 |00:00:00.01 |     556 |  1345K|  1345K|  496K (0)|

|*  4 |     TABLE ACCESS FULL| T        |      1 |  48497 |  1420K|   154   (1)| 00:00:01 |  33334 |00:00:00.01 |     556 |       |       |          |

------------------------------------------------------------------------------------------------------------------------------------------------------

Query Block Name / Object Alias (identified by operation id):

-------------------------------------------------------------

   1 - SEL$C33C846D

   2 - SEL$5771D262 / VW_DAG_0@SEL$C33C846D

   3 - SEL$5771D262

   4 - SEL$5771D262 / T@SEL$1

Predicate Information (identified by operation id):

---------------------------------------------------

   4 - filter(("FLAG1"=1 AND "FLAG2"=1 AND "FLAG3"=1))

Note

-----

   - dynamic statistics used: dynamic sampling (level=2)

   - 1 Sql Plan Directive used for this statement


--//设置OPTIMIZER_ADAPTIVE_STATISTICS=true的情况下,做了动态取样(level=2).产生新的子光标.


SCOTT@test01p> select sql_id,child_number,is_reoptimizable  from v$sql  where sql_id ='872fdta99gdk8';

SQL_ID                CHILD_NUMBER I

------------- -------------------- -

872fdta99gdk8                    0 Y

872fdta99gdk8                    1 N

872fdta99gdk8                    2 N


SCOTT@test01p> @ share 872fdta99gdk8

old  15:           and q.sql_id like ''&1''',

new  15:           and q.sql_id like ''872fdta99gdk8''',

SQL_TEXT                       = select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1

SQL_ID                         = 872fdta99gdk8

ADDRESS                        = 000007FF1393F830

CHILD_ADDRESS                  = 000007FF13D9C198

CHILD_NUMBER                   = 0

USE_FEEDBACK_STATS             = Y

REASON                         = 048Auto Reoptimization Mismatch(1)3x43241945600

--------------------------------------------------

SQL_TEXT                       = select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1

SQL_ID                         = 872fdta99gdk8

ADDRESS                        = 000007FF1393F830

CHILD_ADDRESS                  = 000007FF115A7E58

CHILD_NUMBER                   = 1

REASON                         = 13Optimizer mismatch(12)2x440<_optimizer_dsdir_usage_control> 0                    126                

istics> false

true                <_optimizer_use_feedback_for_join> false                true                <_optimizer_ads_for_pq> false                true                

--------------------------------------------------

SQL_TEXT                       = select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1

SQL_ID                         = 872fdta99gdk8

ADDRESS                        = 000007FF1393F830

CHILD_ADDRESS                  = 000007FF0FDBE618

CHILD_NUMBER                   = 2

OPTIMIZER_MISMATCH             = Y

REASON                         =

--------------------------------------------------


SCOTT@test01p> exec dbms_spd.flush_sql_plan_directive;

PL/SQL procedure successfully completed.


SELECT directive_id

      ,TYPE

      ,enabled

      ,state

      ,notes

      ,reason

  FROM dba_sql_plan_directives

 WHERE directive_id IN (SELECT directive_id

                          FROM dba_sql_plan_dir_objects

                         WHERE owner = USER AND object_name = 'T');


        DIRECTIVE_ID TYPE                 ENA STATE  NOTES                                              REASON

-------------------- -------------------- --- ------ -------------------------------------------------- ------------------------------------

14350253949522184195 DYNAMIC_SAMPLING_RES YES USABLE NEW

                     ULT                             undant>NO{(SCOTT.T, num_rows

                                                     =100000) - (SQL_ID:4k5yrxfcvd5qb, T.CARD=48497[-2

                                                     -2])}


17342821566768621333 DYNAMIC_SAMPLING     YES USABLE MISSING_STATS

                                                     state>NO{EC(SCOTT

                                                     .T)[FLAG1, FLAG2, FLAG3]}


--//多了一行,动态取样分析后估计T.CARD=48497,虽然与实际A-Rows=33334还是存在很大偏差.指导提示是MISSING_STATS.

--//补充说明SQL_ID:4k5yrxfcvd5qb,我没有查询到对于sql语句,有点奇怪!!


SCOTT@test01p> exec dbms_stats.gather_table_stats(user,'T',options=>'gather auto',no_invalidate=>false);

PL/SQL procedure successfully completed.


SCOTT@test01p> select column_name,num_buckets,histogram from user_tab_col_statistics where table_name ='T';

COLUMN_NAME NUM_BUCKETS HISTOGRAM

----------- ----------- ---------------

ID                    1 NONE

NAME                  1 NONE

FLAG1                 3 FREQUENCY

FLAG2                 3 FREQUENCY

FLAG3                 3 FREQUENCY


--//并没有指导建议生成相关列的统计信息.

--//实际上12cR2引入新参数AUTO_STAT_EXTENSIONS控制extended stats的收集,缺省设置off.(没有打开).设置AUTO_STAT_EXTENSIONS=on可以打开.

SCOTT@test01p> select dbms_stats.get_prefs('AUTO_STAT_EXTENSIONS') c10 from dual;

C10

----------

OFF


SCOTT@test01p> exec dbms_stats.set_global_prefs('AUTO_STAT_EXTENSIONS','ON') ;

PL/SQL procedure successfully completed.


SCOTT@test01p> select dbms_stats.get_prefs('AUTO_STAT_EXTENSIONS') c10 from dual;

C10

----------

ON


SCOTT@test01p> exec dbms_stats.gather_table_stats(user,'T',options=>'gather auto',no_invalidate=>false);

PL/SQL procedure successfully completed.


SCOTT@test01p> column COLUMN_NAME format a30

SCOTT@test01p> select column_name,num_buckets,histogram from user_tab_col_statistics where table_name ='T';

COLUMN_NAME                    NUM_BUCKETS HISTOGRAM

------------------------------ ----------- ---------------

ID                                       1 NONE

NAME                                     1 NONE

FLAG1                                    3 FREQUENCY

FLAG2                                    3 FREQUENCY

FLAG3                                    3 FREQUENCY

SYS_STS0SR$HPC$E#KVDPEN#0R2JOU           3 FREQUENCY

6 rows selected.


SCOTT@test01p> column EXTENSION_name format a30

SCOTT@test01p> select * from user_stat_extensions where table_name ='T';

TABLE_NAME EXTENSION_NAME                 EXTENSION                 CREATOR DRO

---------- ------------------------------ ------------------------- ------- ---

T          SYS_STS0SR$HPC$E#KVDPEN#0R2JOU ("FLAG1","FLAG2","FLAG3") SYSTEM  YES


--//可以发现现在收集了相关列("FLAG1","FLAG2","FLAG3")的统计,并且建立了直方图.


SCOTT@test01p> alter session set OPTIMIZER_ADAPTIVE_STATISTICS=false ;

Session altered.


SCOTT@test01p> select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1;

COUNT(DISTINCTNAME)

-------------------

                  1


SCOTT@test01p> @ dpc '' ''

PLAN_TABLE_OUTPUT

-------------------------------------

SQL_ID  872fdta99gdk8, child number 1

-------------------------------------

select count(distinct name) from t where flag1=1 and flag2=1 and flag3=1

Plan hash value: 2359337548

------------------------------------------------------------------------------------------------------------------------------------------------------

| Id  | Operation            | Name     | Starts | E-Rows |E-Bytes| Cost (%CPU)| E-Time   | A-Rows |   A-Time   | Buffers |  OMem |  1Mem | Used-Mem |

------------------------------------------------------------------------------------------------------------------------------------------------------

|   0 | SELECT STATEMENT     |          |      1 |        |       |   156 (100)|          |      1 |00:00:00.01 |     556 |       |       |          |

|   1 |  SORT AGGREGATE      |          |      1 |      1 |    12 |            |          |      1 |00:00:00.01 |     556 |       |       |          |

|   2 |   VIEW               | VW_DAG_0 |      1 |      1 |    12 |   156   (2)| 00:00:01 |      1 |00:00:00.01 |     556 |       |       |          |

|   3 |    HASH GROUP BY     |          |      1 |      1 |    30 |   156   (2)| 00:00:01 |      1 |00:00:00.01 |     556 |  1345K|  1345K|  507K (0)|

|*  4 |     TABLE ACCESS FULL| T        |      1 |  33334 |   976K|   154   (1)| 00:00:01 |  33334 |00:00:00.01 |     556 |       |       |          |

------------------------------------------------------------------------------------------------------------------------------------------------------

Query Block Name / Object Alias (identified by operation id):

-------------------------------------------------------------

   1 - SEL$C33C846D

   2 - SEL$5771D262 / VW_DAG_0@SEL$C33C846D

   3 - SEL$5771D262

   4 - SEL$5771D262 / T@SEL$1

Predicate Information (identified by operation id):

---------------------------------------------------

   4 - filter(("FLAG1"=1 AND "FLAG2"=1 AND "FLAG3"=1))


--//可以发现E-Rows已经正确修正.


SCOTT@test01p> exec dbms_spd.flush_sql_plan_directive;

PL/SQL procedure successfully completed.



        DIRECTIVE_ID TYPE                 ENA STATE                NOTES                                              REASON

-------------------- -------------------- --- -------------------- -------------------------------------------------- ------------------------------------

14350253949522184195 DYNAMIC_SAMPLING_RES YES USABLE               NEW

                     ULT                                           undant>NO{(SCOTT.T, num_rows

                                                                   =100000) - (SQL_ID:4k5yrxfcvd5qb, T.CARD=48497[-2

                                                                   -2])}


17342821566768621333 DYNAMIC_SAMPLING     YES SUPERSEDED           HAS_STATS

                                                                   e>NO{EC(SCOTT.T)[

                                                                   FLAG1, FLAG2, FLAG3]}


--//注意看现在不是MISSING_STATS而是提示HAS_STATS.   SUPERSEDED 表示 取代,接替.

--//有了相关列统计其它涉及相关列的查询就不会在动态取样,而是估计行数与实际行数接近.而且执行其它类似语句也不会出现is_reoptimizable='Y'的情况.


SCOTT@test01p> select max(id) from t where flag1=1 and flag2=1 and flag3=1;

   MAX(ID)

----------

    100000


SCOTT@test01p> @ dpc '' ''

PLAN_TABLE_OUTPUT

-------------------------------------

SQL_ID  6stmvx0gcybbg, child number 0

-------------------------------------

select max(id) from t where flag1=1 and flag2=1 and flag3=1

Plan hash value: 2966233522

---------------------------------------------------------------------------------------------------------------------

| Id  | Operation          | Name | Starts | E-Rows |E-Bytes| Cost (%CPU)| E-Time   | A-Rows |   A-Time   | Buffers |

---------------------------------------------------------------------------------------------------------------------

|   0 | SELECT STATEMENT   |      |      1 |        |       |   154 (100)|          |      1 |00:00:00.01 |     556 |

|   1 |  SORT AGGREGATE    |      |      1 |      1 |    14 |            |          |      1 |00:00:00.01 |     556 |

|*  2 |   TABLE ACCESS FULL| T    |      1 |  33334 |   455K|   154   (1)| 00:00:01 |  33334 |00:00:00.01 |     556 |

---------------------------------------------------------------------------------------------------------------------

Query Block Name / Object Alias (identified by operation id):

-------------------------------------------------------------

   1 - SEL$1

   2 - SEL$1 / T@SEL$1

Predicate Information (identified by operation id):

---------------------------------------------------

   2 - filter(("FLAG1"=1 AND "FLAG2"=1 AND "FLAG3"=1))


SCOTT@test01p> select sql_id,child_number,is_reoptimizable  from v$sql  where sql_id ='6stmvx0gcybbg';

SQL_ID        CHILD_NUMBER I

------------- ------------ -

6stmvx0gcybbg            0 N


--//is_reoptimizable = 'N'.


总结:

--//12cR2做了一些改进,optimizer_adaptive_statistics=false,避免大量的动态取样对性能的影响.另外即使设置optimizer_adaptive_statistics=true.

--//oracle也保存了动态取样的结果.

--//dbms_stats引入新的参数AUTO_STAT_EXTENSIONS,缺省是off.设置on后再分析自动建立扩展统计信息.


来自 “ ITPUB博客 ” ,链接:http://blog.itpub.net/267265/viewspace-2286394/,如需转载,请注明出处,否则将追究法律责任。

转载于:http://blog.itpub.net/267265/viewspace-2286394/

你可能感兴趣的:([20181225]12CR2 SQL Plan Directives.txt)