12條語句學會oracle cbo計算(六)
12條語句學會oracle cbo計算(六)
工作中,你可能會遇到統計資料正確,但就是不走你想要的執行計劃的情況,最後一般透過hint或sql_profile解決.一條sql語句的效能主要依賴於好的物理結構,準確的系統統計資料,準確的物件統計資料,合理的查詢最佳化器引數,合理的系統引數.這些因素也就是cost計算基本引數.因此產生的sql最佳化技術有很多,大概包含:改變訪問結構,修改sql語句,加hint,改變執行環境,sql profile,stored outlines,sql plan baseline.其實無外乎就是透過調整sql適應環境或者透過調整環境適應sql.如果清楚Cost值的演算法,就可以透過演算法的對比,找到問題的真正原因,更有針對性的去解決問題.
全文主要參考Jonathan Lewis的<>和黃瑋(fuyuncat)的<>,特別黃瑋(fuyuncat)的這本,是非常值得去學習的.
準備用14篇來描述完,前2篇是統計資料,演算法公式說明,後12篇用12條語句分別去套用說明.
本篇例子的特徵是單表,全表掃描,條件值常量,無直方圖,單條件,排序,和上一篇差別是排序
--產生測試資料
drop table scott.t_test1 purge;
create table scott.t_test1 as select * from dba_objects;
begin
dbms_stats.gather_table_stats('scott','t_test1');
end;
--產生語句的執行計劃
--這裡我是在pl/sql developer,是因為不用象10053那麼麻煩就可以產生想要的幾個值用以對比.
explain plan for select * from scott.t_test1 where owner='SYSTEM' order by object_id;
SELECT lpad(' ', 2 * (LEVEL - 1)) || operation operation,
options,
object_name,
cardinality,
bytes,
io_cost,
cpu_cost,
cost,
time
FROM plan_table
START WITH id = 0
CONNECT BY PRIOR id = parent_id;
/*
OPERATION OPTIONS OBJECT_NAME CARDINALITY BYTES IO_COST CPU_COST COST TIME
SELECT STATEMENT 3754 367892 430 67549918 432 6
SORT ORDER BY 3754 367892 430 67549918 432 6
TABLE ACCESS FULL T_TEST1 3754 367892 343 27298256 344 5
*/
--查詢表的統計資料
select rpad(table_name, 10, ' ') table_name,
rpad(num_rows, 10, ' ') num_rows,
rpad(blocks, 10, ' ') blocks,
avg_row_len
from dba_tables
where owner = 'SCOTT'
and table_name = 'T_TEST1';
/*
TABLE_NAME NUM_ROWS BLOCKS AVG_ROW_LEN
T_TEST1 86335 1261 98
*/
--查詢列的統計資料
select rpad(column_name, 12, ' ') column_name,
rpad(num_distinct, 8, ' ') num_distinct,
rpad(utl_raw.cast_to_varchar2(low_value), 15, ' ') low_value,
rpad(utl_raw.cast_to_varchar2(high_value), 10, ' ') high_value,
rpad(nullable, 8, ' ') nullable,
rpad(num_nulls, 8, ' ') num_nulls,
rpad(avg_col_len, 6, ' ') avg_col_len,
rpad(density, 20, ' ') density,
histogram
from dba_tab_columns
where owner = 'SCOTT'
and table_name = 'T_TEST1'
and column_name ='OWNER';
/*
COLUMN_NAME NUM_DISTINCT LOW_VALUE HIGH_VALUE NULLABLE NUM_NULLS AVG_COL_LEN DENSITY HISTOGRAM
OWNER 23 APEX_030200 XDB Y 0 6 .0434782608695652 NONE
*/
--查詢各列平均長度之和
select sum(avg_col_len),count(1)
from dba_tab_col_statistics
where owner='SCOTT' and table_name = 'T_TEST1'
/*
SUM(AVG_COL_LEN) COUNT(1)
100 15
*/
--查詢最佳化器引數
select rpad(name,40,' ') name,rpad(value,20,' ') value,isdefault
from (select nam.ksppinm name,
val.KSPPSTVL value,
--nam.ksppdesc description,
val.ksppstdf isdefault
from sys.x$ksppi nam, sys.x$ksppcv val
where nam.inst_id = val.inst_id
and nam.indx = val.indx)
where name in
('_db_file_optimizer_read_count', 'db_file_multiblock_read_count',
'_optimizer_block_size', '_table_scan_cost_plus_one',
'_optimizer_ceil_cost', '_optimizer_cost_model',
'_optimizer_cache_stats', '_smm_auto_min_io_size',
'_smm_auto_max_io_size', '_smm_min_size', '_smm_max_size',
'_smm_px_max_size', 'sort_area_retained_size', 'sort_area_size',
'workarea_size_policy','_optimizer_percent_parallel');
/*
NAME VALUE ISDEFAULT
db_file_multiblock_read_count 116 TRUE
_db_file_optimizer_read_count 8 TRUE
sort_area_size 65536 TRUE
sort_area_retained_size 0 TRUE
_optimizer_cost_model CHOOSE TRUE
_optimizer_cache_stats FALSE TRUE
_table_scan_cost_plus_one TRUE TRUE
workarea_size_policy AUTO TRUE
_smm_auto_min_io_size 56 TRUE
_smm_auto_max_io_size 248 TRUE
_smm_min_size 286 TRUE
_smm_max_size 57344 TRUE
_smm_px_max_size 143360 TRUE
_optimizer_percent_parallel 101 TRUE
_optimizer_block_size 8192 TRUE
_optimizer_ceil_cost TRUE TRUE
*/
--查詢系統統計資料
select rpad(pname, '20', ' ') pname,
rpad(pval1, '20', ' ') pval1,
rpad(pval2, '20', ' ') pval2
from SYS.AUX_STATS$
where sname = 'SYSSTATS_MAIN';
/*
PNAME PVAL1 PVAL2
CPUSPEED
CPUSPEEDNW 3074.07407407407
IOSEEKTIM 10
IOTFRSPEED 4096
MAXTHR
MBRC
MREADTIM
SLAVETHR
SREADTIM
*/
--需要應用第二篇中的公式:
(1)NDV=dba_tab_co1umns.num_distinct
(2)DENS=dba_tab_co1umns.DENSITY
(3)ALLROWS=dba_tab1es.NUM_ROWS
(6)COLNB=dba_tab_co1umns.NULLABLE
(8)ARL=dba_tables.AVG_ROW_LEN
(11)MBRC=最佳化器系統引數_db_fi1e_optimizer_read_count
(12)SAMINIO=最佳化器系統引數_smm_auto_min_io_size*1024
(14)OPTBLKSIZE=最佳化器系統引數_optimizer_b1ock_size
(16)MBDRC=SAMINIO/OPTBLKSIZE
(21)CPUSPEED=系統統計資料CPUSPEEDNW
(22)IOTFRSPEED=系統統計資料IOTFRSPEED
(23)IOSEEKTIM=系統統計資料IOSEEKTIM
(24)SREADTIM = IOSEEKTIM + OPTBLKSIZ/IOTFRSPEED
(26)SAMSIZE=最佳化器系統引數_smm_max_size*1024
(27)SASIZE=最佳化器系統引數_smm_min_size*1024
(29)=的選擇率為: GREATEST(1/NDV,DENS)*DECODE(COLNB= Y,1,NNV/ALLROWS)
(72)IOCOST = (#BLKS/MBRC)*(IOSEEKTIM + MBRC*OPTBLKSIZE/IOTFRSPEED)/(IOSEEKTIM+OPTBLKSIZE/IOTFRSPEED)
(73)CPUCOST = #CPUCYCLES /(CPUSPEED*SREADTIM)/1000
(80)SDSIZE = SROWNUM*SROWSIZE
(81)RROWSIZE = LEAST(ARL, SUM(ACL1~n))
(82)SROWSIZE = RROWSIZE + 10 + CEIL(RROWSIZE/10)
(83)如果SDSIZE > SASIZE,說明排序需要寫入磁碟,否則IOCOST[Sort]=0
(84)SORTWIDTH = FLOOR((SAMSIZE-((60*SAMSIZE/1024/320-40*(SAMSIZE/1024/320-1))+LOG(2,MBDRC)*80)*1024)/((SAMINIO+OPTBLKSIZE)*2.5))
(85)INTRUNS = GREATEST(CEIL(SDSIZE/SAMSIZE), 2)
(86)MERGES = CEIL(LOG(SORTWIDTH,INIRUNS))
(87)SORTBLKS = CEIL(SDSIZE/(OPTBLKSIZE-24))
(88)PASSIO = CEIL(SORTBLKS*(MBDRC*MREADTIM/SREADTIM)/(MBDRC+1 )/(MBRC-1 ))*2 + CEIL(S0RTBLKS*(MBRC-1-MBDRC)/(MBDRC+1)/(MBRC-1))*2
(89)IOCOST[Sort]= SORTBLKS + (PASSIO * MERGES))
--套用上面的公式及資料進行計算
表名:T_TEST1
(3)ALLROWS=dba_tab1es.NUM_ROWS=86335
條件列名:OWNER
(1)NDV=dba_tab_co1umns.num_distinct=23
(2)DENS=dba_tab_co1umns.DENSITY=.0434782608695652
(6)COLNB=dba_tab_co1umns.NULLABLE=Y
(8)ARL=dba_tables.AVG_ROW_LEN=98
(11)MBRC=最佳化器系統引數_db_fi1e_optimizer_read_count=8
(12)SAMINIO=最佳化器系統引數_smm_auto_min_io_size*1024=56*1024=57344
(14)OPTBLKSIZE=最佳化器系統引數_optimizer_b1ock_size=8192
(16)MBDRC=SAMINIO/OPTBLKSIZE=57344/8192=7
(21)CPUSPEED=系統統計資料CPUSPEEDNW=3074.07407407407
(22)IOTFRSPEED=系統統計資料IOTFRSPEED=4096
(23)IOSEEKTIM=系統統計資料IOSEEKTIM=10
(24)SREADTIM = IOSEEKTIM + OPTBLKSIZ/IOTFRSPEED=10+8192/4096=12
(26)SAMSIZE=最佳化器系統引數_smm_max_size*1024=58720256
(27)SASIZE=最佳化器系統引數_smm_min_size*1024=292864
(29)=的選擇率為: GREATHST(1/NDV,DENS)*DECODE(COLNB,Y,1,NNV/ALLROWS)
SEL=GREATEST(1/23,.0434782608695652)*DECODE('Y',Y,1,86335/86335)
=0.0434782608695652
ROWS=ALLROWS*SEL=86335*0.0434782608695652=3753.69565217391=3754
--TABLE ACCESS FULL
(72)IOCOST = (#BLKS/MBRC)*(IOSEEKTIM + MBRC*OPTBLKSIZE/IOTFRSPEED)/(IOSEEKTIM+OPTBLKSIZE/IOTFRSPEED)
= (1261/8)*(10 + 8*8192/4096)/(10 + 8192/4096)=341.520833333333
由於_optimizer_ceil_cost=true,_table_scan_cost_plus_one=true,所以微調為:
IOCOST=ceil(341.520833333333)+1=343
(73)CPUCOST = #CPUCYCLES /(CPUSPEED*SREADTIM)/1000
=27298256/(3074.07407407407*12)/1000
=0.740012963855423
COST=IOCOST+CPUCOST=343+0.740012963855423=343.740012963855423=344
--SORT ORDER BY
(81)RROWSIZE = LEAST(ARL, SUM(ACL1~n))
=LEAST(98, 100)
=98
(82)SROWSIZE = RROWSIZE + 10 + CEIL(RROWSIZE/10)
=98+10+CEIL(98/10)
=118
SROWNUM=ROWS
(80)SDSIZE = SROWNUM*SROWSIZE
=3754*118
=442972
(83)如果SDSIZE > SASIZE,說明排序需要寫入磁碟,否則IOCOST[Sort]=0
SDSIZE(442972)>SASIZE(292864)
(84)SORTWIDTH = FLOOR((SAMSIZE-((60*SAMSIZE/1024/320-40*(SAMSIZE/1024/320-1))+LOG(2,MBDRC)*80)*1024)/((SAMINIO+OPTBLKSIZE)*2.5))
=FLOOR((58720256-((60*58720256/1024/320-40*(58720256/1024/320-1))+LOG(2,7)*80)*1024)/((57344+8192)*2.5))
=334
(85)INTRUNS = GREATEST(CEIL(SDSIZE/SAMSIZE), 2)
=GREATEST(CEIL(442972/58720256), 2)
=2
(86)MERGES = CEIL(LOG(SORTWIDTH,INIRUNS))
=CEIL(LOG(334,2))
=1
(87)SORTBLKS = CEIL(SDSIZE/(OPTBLKSIZE-24))
=CEIL(442972/(8192-24))
=55
(88)PASSIO = CEIL(SORTBLKS*(MBDRC*MREADTIM/SREADTIM)/(MBDRC+1 )/(MBRC-1 ))*2 + CEIL(S0RTBLKS*(MBRC-1-MBDRC)/(MBDRC+1)/(MBRC-1))*2
=CEIL(55*(7*26/12)/(7+1 )/(8-1 ))*2 + CEIL(55*(8-1-7)/(7+1)/(8-1))*2
=30
(89)IOCOST[Sort]= SORTBLKS + (PASSIO * MERGES)
= 55 + (30 * 1)
=85
IOCOST=IOCOST[TABLE ACCESS FULL]+IOCOST[Sort]=343+85=428
CPUCOST=#CPUCYCLES /(CPUSPEED*SREADTIM)/1000
= 67549918/(3074.07407407407*12)/1000
=1.83117247590362
COST= IOCOST+CPUCOST=428+1.83117247590362=429.83117247590362=430
--可以看到,結果與執行計劃基本相同.
ROWS=ALLROWS*SEL=86335*0.0434782608695652=3753.69565217391=3754
IOCOST[TABLE ACCESS FULL]=ceil(341.520833333333)+1=343
IOCOST[Sort]=55 + (30 * 1)=85
IOCOST=IOCOST[TABLE ACCESS FULL]+IOCOST[Sort]=343+85=428
CPUCOST = 67549918/(3074.07407407407*12)/1000 =1.83117247590362
COST=IOCOST+CPUCOST=428+1.83117247590362=429.83117247590362=430
--從排序成本的計算,可以看出,排序主要所消耗的成本在IO成本和決定排序成本的幾個關鍵引數.
工作中,你可能會遇到統計資料正確,但就是不走你想要的執行計劃的情況,最後一般透過hint或sql_profile解決.一條sql語句的效能主要依賴於好的物理結構,準確的系統統計資料,準確的物件統計資料,合理的查詢最佳化器引數,合理的系統引數.這些因素也就是cost計算基本引數.因此產生的sql最佳化技術有很多,大概包含:改變訪問結構,修改sql語句,加hint,改變執行環境,sql profile,stored outlines,sql plan baseline.其實無外乎就是透過調整sql適應環境或者透過調整環境適應sql.如果清楚Cost值的演算法,就可以透過演算法的對比,找到問題的真正原因,更有針對性的去解決問題.
全文主要參考Jonathan Lewis的<>和黃瑋(fuyuncat)的<
準備用14篇來描述完,前2篇是統計資料,演算法公式說明,後12篇用12條語句分別去套用說明.
本篇例子的特徵是單表,全表掃描,條件值常量,無直方圖,單條件,排序,和上一篇差別是排序
--產生測試資料
drop table scott.t_test1 purge;
create table scott.t_test1 as select * from dba_objects;
begin
dbms_stats.gather_table_stats('scott','t_test1');
end;
--產生語句的執行計劃
--這裡我是在pl/sql developer,是因為不用象10053那麼麻煩就可以產生想要的幾個值用以對比.
explain plan for select * from scott.t_test1 where owner='SYSTEM' order by object_id;
SELECT lpad(' ', 2 * (LEVEL - 1)) || operation operation,
options,
object_name,
cardinality,
bytes,
io_cost,
cpu_cost,
cost,
time
FROM plan_table
START WITH id = 0
CONNECT BY PRIOR id = parent_id;
/*
OPERATION OPTIONS OBJECT_NAME CARDINALITY BYTES IO_COST CPU_COST COST TIME
SELECT STATEMENT 3754 367892 430 67549918 432 6
SORT ORDER BY 3754 367892 430 67549918 432 6
TABLE ACCESS FULL T_TEST1 3754 367892 343 27298256 344 5
*/
--查詢表的統計資料
select rpad(table_name, 10, ' ') table_name,
rpad(num_rows, 10, ' ') num_rows,
rpad(blocks, 10, ' ') blocks,
avg_row_len
from dba_tables
where owner = 'SCOTT'
and table_name = 'T_TEST1';
/*
TABLE_NAME NUM_ROWS BLOCKS AVG_ROW_LEN
T_TEST1 86335 1261 98
*/
--查詢列的統計資料
select rpad(column_name, 12, ' ') column_name,
rpad(num_distinct, 8, ' ') num_distinct,
rpad(utl_raw.cast_to_varchar2(low_value), 15, ' ') low_value,
rpad(utl_raw.cast_to_varchar2(high_value), 10, ' ') high_value,
rpad(nullable, 8, ' ') nullable,
rpad(num_nulls, 8, ' ') num_nulls,
rpad(avg_col_len, 6, ' ') avg_col_len,
rpad(density, 20, ' ') density,
histogram
from dba_tab_columns
where owner = 'SCOTT'
and table_name = 'T_TEST1'
and column_name ='OWNER';
/*
COLUMN_NAME NUM_DISTINCT LOW_VALUE HIGH_VALUE NULLABLE NUM_NULLS AVG_COL_LEN DENSITY HISTOGRAM
OWNER 23 APEX_030200 XDB Y 0 6 .0434782608695652 NONE
*/
--查詢各列平均長度之和
select sum(avg_col_len),count(1)
from dba_tab_col_statistics
where owner='SCOTT' and table_name = 'T_TEST1'
/*
SUM(AVG_COL_LEN) COUNT(1)
100 15
*/
--查詢最佳化器引數
select rpad(name,40,' ') name,rpad(value,20,' ') value,isdefault
from (select nam.ksppinm name,
val.KSPPSTVL value,
--nam.ksppdesc description,
val.ksppstdf isdefault
from sys.x$ksppi nam, sys.x$ksppcv val
where nam.inst_id = val.inst_id
and nam.indx = val.indx)
where name in
('_db_file_optimizer_read_count', 'db_file_multiblock_read_count',
'_optimizer_block_size', '_table_scan_cost_plus_one',
'_optimizer_ceil_cost', '_optimizer_cost_model',
'_optimizer_cache_stats', '_smm_auto_min_io_size',
'_smm_auto_max_io_size', '_smm_min_size', '_smm_max_size',
'_smm_px_max_size', 'sort_area_retained_size', 'sort_area_size',
'workarea_size_policy','_optimizer_percent_parallel');
/*
NAME VALUE ISDEFAULT
db_file_multiblock_read_count 116 TRUE
_db_file_optimizer_read_count 8 TRUE
sort_area_size 65536 TRUE
sort_area_retained_size 0 TRUE
_optimizer_cost_model CHOOSE TRUE
_optimizer_cache_stats FALSE TRUE
_table_scan_cost_plus_one TRUE TRUE
workarea_size_policy AUTO TRUE
_smm_auto_min_io_size 56 TRUE
_smm_auto_max_io_size 248 TRUE
_smm_min_size 286 TRUE
_smm_max_size 57344 TRUE
_smm_px_max_size 143360 TRUE
_optimizer_percent_parallel 101 TRUE
_optimizer_block_size 8192 TRUE
_optimizer_ceil_cost TRUE TRUE
*/
--查詢系統統計資料
select rpad(pname, '20', ' ') pname,
rpad(pval1, '20', ' ') pval1,
rpad(pval2, '20', ' ') pval2
from SYS.AUX_STATS$
where sname = 'SYSSTATS_MAIN';
/*
PNAME PVAL1 PVAL2
CPUSPEED
CPUSPEEDNW 3074.07407407407
IOSEEKTIM 10
IOTFRSPEED 4096
MAXTHR
MBRC
MREADTIM
SLAVETHR
SREADTIM
*/
--需要應用第二篇中的公式:
(1)NDV=dba_tab_co1umns.num_distinct
(2)DENS=dba_tab_co1umns.DENSITY
(3)ALLROWS=dba_tab1es.NUM_ROWS
(6)COLNB=dba_tab_co1umns.NULLABLE
(8)ARL=dba_tables.AVG_ROW_LEN
(11)MBRC=最佳化器系統引數_db_fi1e_optimizer_read_count
(12)SAMINIO=最佳化器系統引數_smm_auto_min_io_size*1024
(14)OPTBLKSIZE=最佳化器系統引數_optimizer_b1ock_size
(16)MBDRC=SAMINIO/OPTBLKSIZE
(21)CPUSPEED=系統統計資料CPUSPEEDNW
(22)IOTFRSPEED=系統統計資料IOTFRSPEED
(23)IOSEEKTIM=系統統計資料IOSEEKTIM
(24)SREADTIM = IOSEEKTIM + OPTBLKSIZ/IOTFRSPEED
(26)SAMSIZE=最佳化器系統引數_smm_max_size*1024
(27)SASIZE=最佳化器系統引數_smm_min_size*1024
(29)=的選擇率為: GREATEST(1/NDV,DENS)*DECODE(COLNB= Y,1,NNV/ALLROWS)
(72)IOCOST = (#BLKS/MBRC)*(IOSEEKTIM + MBRC*OPTBLKSIZE/IOTFRSPEED)/(IOSEEKTIM+OPTBLKSIZE/IOTFRSPEED)
(73)CPUCOST = #CPUCYCLES /(CPUSPEED*SREADTIM)/1000
(80)SDSIZE = SROWNUM*SROWSIZE
(81)RROWSIZE = LEAST(ARL, SUM(ACL1~n))
(82)SROWSIZE = RROWSIZE + 10 + CEIL(RROWSIZE/10)
(83)如果SDSIZE > SASIZE,說明排序需要寫入磁碟,否則IOCOST[Sort]=0
(84)SORTWIDTH = FLOOR((SAMSIZE-((60*SAMSIZE/1024/320-40*(SAMSIZE/1024/320-1))+LOG(2,MBDRC)*80)*1024)/((SAMINIO+OPTBLKSIZE)*2.5))
(85)INTRUNS = GREATEST(CEIL(SDSIZE/SAMSIZE), 2)
(86)MERGES = CEIL(LOG(SORTWIDTH,INIRUNS))
(87)SORTBLKS = CEIL(SDSIZE/(OPTBLKSIZE-24))
(88)PASSIO = CEIL(SORTBLKS*(MBDRC*MREADTIM/SREADTIM)/(MBDRC+1 )/(MBRC-1 ))*2 + CEIL(S0RTBLKS*(MBRC-1-MBDRC)/(MBDRC+1)/(MBRC-1))*2
(89)IOCOST[Sort]= SORTBLKS + (PASSIO * MERGES))
--套用上面的公式及資料進行計算
表名:T_TEST1
(3)ALLROWS=dba_tab1es.NUM_ROWS=86335
條件列名:OWNER
(1)NDV=dba_tab_co1umns.num_distinct=23
(2)DENS=dba_tab_co1umns.DENSITY=.0434782608695652
(6)COLNB=dba_tab_co1umns.NULLABLE=Y
(8)ARL=dba_tables.AVG_ROW_LEN=98
(11)MBRC=最佳化器系統引數_db_fi1e_optimizer_read_count=8
(12)SAMINIO=最佳化器系統引數_smm_auto_min_io_size*1024=56*1024=57344
(14)OPTBLKSIZE=最佳化器系統引數_optimizer_b1ock_size=8192
(16)MBDRC=SAMINIO/OPTBLKSIZE=57344/8192=7
(21)CPUSPEED=系統統計資料CPUSPEEDNW=3074.07407407407
(22)IOTFRSPEED=系統統計資料IOTFRSPEED=4096
(23)IOSEEKTIM=系統統計資料IOSEEKTIM=10
(24)SREADTIM = IOSEEKTIM + OPTBLKSIZ/IOTFRSPEED=10+8192/4096=12
(26)SAMSIZE=最佳化器系統引數_smm_max_size*1024=58720256
(27)SASIZE=最佳化器系統引數_smm_min_size*1024=292864
(29)=的選擇率為: GREATHST(1/NDV,DENS)*DECODE(COLNB,Y,1,NNV/ALLROWS)
SEL=GREATEST(1/23,.0434782608695652)*DECODE('Y',Y,1,86335/86335)
=0.0434782608695652
ROWS=ALLROWS*SEL=86335*0.0434782608695652=3753.69565217391=3754
--TABLE ACCESS FULL
(72)IOCOST = (#BLKS/MBRC)*(IOSEEKTIM + MBRC*OPTBLKSIZE/IOTFRSPEED)/(IOSEEKTIM+OPTBLKSIZE/IOTFRSPEED)
= (1261/8)*(10 + 8*8192/4096)/(10 + 8192/4096)=341.520833333333
由於_optimizer_ceil_cost=true,_table_scan_cost_plus_one=true,所以微調為:
IOCOST=ceil(341.520833333333)+1=343
(73)CPUCOST = #CPUCYCLES /(CPUSPEED*SREADTIM)/1000
=27298256/(3074.07407407407*12)/1000
=0.740012963855423
COST=IOCOST+CPUCOST=343+0.740012963855423=343.740012963855423=344
--SORT ORDER BY
(81)RROWSIZE = LEAST(ARL, SUM(ACL1~n))
=LEAST(98, 100)
=98
(82)SROWSIZE = RROWSIZE + 10 + CEIL(RROWSIZE/10)
=98+10+CEIL(98/10)
=118
SROWNUM=ROWS
(80)SDSIZE = SROWNUM*SROWSIZE
=3754*118
=442972
(83)如果SDSIZE > SASIZE,說明排序需要寫入磁碟,否則IOCOST[Sort]=0
SDSIZE(442972)>SASIZE(292864)
(84)SORTWIDTH = FLOOR((SAMSIZE-((60*SAMSIZE/1024/320-40*(SAMSIZE/1024/320-1))+LOG(2,MBDRC)*80)*1024)/((SAMINIO+OPTBLKSIZE)*2.5))
=FLOOR((58720256-((60*58720256/1024/320-40*(58720256/1024/320-1))+LOG(2,7)*80)*1024)/((57344+8192)*2.5))
=334
(85)INTRUNS = GREATEST(CEIL(SDSIZE/SAMSIZE), 2)
=GREATEST(CEIL(442972/58720256), 2)
=2
(86)MERGES = CEIL(LOG(SORTWIDTH,INIRUNS))
=CEIL(LOG(334,2))
=1
(87)SORTBLKS = CEIL(SDSIZE/(OPTBLKSIZE-24))
=CEIL(442972/(8192-24))
=55
(88)PASSIO = CEIL(SORTBLKS*(MBDRC*MREADTIM/SREADTIM)/(MBDRC+1 )/(MBRC-1 ))*2 + CEIL(S0RTBLKS*(MBRC-1-MBDRC)/(MBDRC+1)/(MBRC-1))*2
=CEIL(55*(7*26/12)/(7+1 )/(8-1 ))*2 + CEIL(55*(8-1-7)/(7+1)/(8-1))*2
=30
(89)IOCOST[Sort]= SORTBLKS + (PASSIO * MERGES)
= 55 + (30 * 1)
=85
IOCOST=IOCOST[TABLE ACCESS FULL]+IOCOST[Sort]=343+85=428
CPUCOST=#CPUCYCLES /(CPUSPEED*SREADTIM)/1000
= 67549918/(3074.07407407407*12)/1000
=1.83117247590362
COST= IOCOST+CPUCOST=428+1.83117247590362=429.83117247590362=430
--可以看到,結果與執行計劃基本相同.
ROWS=ALLROWS*SEL=86335*0.0434782608695652=3753.69565217391=3754
IOCOST[TABLE ACCESS FULL]=ceil(341.520833333333)+1=343
IOCOST[Sort]=55 + (30 * 1)=85
IOCOST=IOCOST[TABLE ACCESS FULL]+IOCOST[Sort]=343+85=428
CPUCOST = 67549918/(3074.07407407407*12)/1000 =1.83117247590362
COST=IOCOST+CPUCOST=428+1.83117247590362=429.83117247590362=430
--從排序成本的計算,可以看出,排序主要所消耗的成本在IO成本和決定排序成本的幾個關鍵引數.
來自 “ ITPUB部落格 ” ，連結：http://blog.itpub.net/30255758/viewspace-1666852/，如需轉載，請註明出處，否則將追究法律責任。
相關文章
- 12條語句學會oracle cbo計算(七)Oracle
- 12條語句學會oracle cbo計算(五)Oracle
- 12條語句學會oracle cbo計算(四)Oracle
- 12條語句學會oracle cbo計算(三)Oracle
- 12條語句學會oracle cbo計算(二)Oracle
- 12條語句學會oracle cbo計算(一)Oracle
- 12條語句學會oracle cbo計算(十四)Oracle
- 12條語句學會oracle cbo計算(十三)Oracle
- 12條語句學會oracle cbo計算(十一)Oracle
- 12條語句學會oracle cbo計算(十)Oracle
- 12條語句學會oracle cbo計算(九)Oracle
- 12條語句學會oracle cbo計算(十二)Oracle
- 【cbo計算公式】Join 選擇率(六)公式
- oracle中的條件語句Oracle
- 學習Rust 條件語句Rust
- Oracle分頁查詢語句（六）Oracle
- oracle基於cbo成本計算方式說明Oracle
- 【cbo計算公式】CBO基本概念(一)公式
- CBO成本計算初探
- 草根學Python（五） 條件語句和迴圈語句Python
- 條件語句
- Java 數學運算與條件語句全解析Java
- Python學習筆記3（條件語句+迴圈語句）Python筆記
- Python學習筆記（三）——條件語句、迴圈語句Python筆記
- shell程式設計之條件語句程式設計
- 實現四則運算的一條sql語句SQL
- Go 條件語句 - Go 學習記錄Go
- GO 條件語句Go
- javaScript條件語句JavaScript
- 02 shell程式設計之條件語句程式設計
- Oracle 使用一條insert語句完成多表插入Oracle
- day2-計算符和if語句
- 9，javase程式碼實戰-條件語句——switch語句根據使用者消費金額計算折扣(四)Java
- 【cbo計算公式】No Bind Peeking(五)公式
- Swift學習筆記（二十七）——條件語句和Switch-case語句Swift筆記
- Go：條件控制語句Go
- if條件語句sed命令
- 04.條件語句