{"id":809,"date":"2016-07-24T23:41:02","date_gmt":"2016-07-25T02:41:02","guid":{"rendered":"http:\/\/dbtimewizard.com.br\/blog\/?p=809"},"modified":"2019-11-11T18:55:11","modified_gmt":"2019-11-11T21:55:11","slug":"exadata-sera-que-podemos-eliminar-os-indices-das-tabelas","status":"publish","type":"post","link":"https:\/\/dbtimewizard.com.br\/blog\/exadata-sera-que-podemos-eliminar-os-indices-das-tabelas\/","title":{"rendered":"Exadata: Ser\u00e1 que podemos eliminar os \u00edndices das tabelas?"},"content":{"rendered":"

\u00c9 comum ouvir outras pessoas dizendo que no Exadata n\u00e3o precisamos mais dos \u00edndices de banco de dados tradicionais pois o Exadata responde melhor sem eles e que podemos excluir todos os \u00edndices para liberar espa\u00e7o no Storage.<\/span>
\n Na pr\u00e1tica as coisas s\u00e3o um pouco diferentes, existem muitos cen\u00e1rios onde os \u00edndices de banco de dados ainda s\u00e3o cr\u00edticos para garantir o desempenho do banco em um ambiente Exadata.<\/span><\/p>\n

Neste artigo vamos demonstrar duas situa\u00e7\u00f5es onde a utiliza\u00e7\u00e3o do \u00edndice tradicional do banco continua fazendo toda a diferen\u00e7a em termos de performace e at\u00e9 mesmo os recursos fant\u00e1sticos do Exadata n\u00e3o conseguem supera-lo.<\/span><\/p>\n

Utiliza\u00e7\u00e3o de fun\u00e7\u00f5es na consulta<\/span><\/h2>\n

O banco de dados Oracle disp\u00f5e de varias fun\u00e7\u00f5es nativas que podem ser utilizadas na instru\u00e7\u00e3o SQL, muitas dessas fun\u00e7\u00f5es s\u00e3o resolvidas no Exadata Storage que na terminologia do Exadata corresponde ao “cell offloaded”, porem algumas fun\u00e7\u00f5es n\u00e3o est\u00e3o dispon\u00edveis ainda para “cell offloaded” na vers\u00e3o 11.2.0.4, para identificar quais fun\u00e7\u00f5es n\u00e3o s\u00e3o resolvidas pelo Exadata Storage consulte a vis\u00e3o “V$SQLFN_METADATA<\/span>“.<\/span><\/p>\n

 <\/p>\n

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
<\/div><\/td>
SQL><\/span>select NAME,<\/span> OFFLOADABLE from v$sqlfn_metadata<\/span> where name in (<\/span>'MIN'<\/span>,<\/span>'MAX'<\/span>,<\/span>'AVG'<\/span>,<\/span>'SUM'<\/span>,<\/span>'COUNT'<\/span>)<\/span> order by 1<\/span>;<\/span>
\n
\nNAME       OFFLOADABLE
\n----------<\/span> ---------------<\/span>
\nAVG        NO
\nAVG        NO
\nAVG        NO
\nAVG        NO
\nAVG        NO
\nAVG        NO
\nCOUNT<\/span><\/a>      NO
\nCOUNT<\/span><\/a>      NO
\nMAX<\/span><\/a>        NO
\nMAX<\/span><\/a>        NO
\nMIN<\/span><\/a>        NO
\nMIN<\/span><\/a>        NO
\nSUM        NO
\nSUM        NO
\nSUM        NO
\nSUM        NO
\nSUM        NO
\nSUM        NO
\n
\n18<\/span> rows selected.<\/span>
\n
\nSQL><\/span><\/div><\/td><\/tr><\/tbody><\/table><\/div>\n

 <\/p>\n

Uma das situa\u00e7\u00f5es em que o tradicional \u00edndice de\u00a0banco de dados pode ajudar a melhorar a performace no Exadata s\u00e3o as instru\u00e7\u00f5es SQL que utilizam fun\u00e7\u00f5es que n\u00e3o s\u00e3o resolvidas no “cell offloaded”. Para demonstrar essa situa\u00e7\u00e3o vamos executar uma consulta com a fun\u00e7\u00e3o MIN() sem a utiliza\u00e7\u00e3o de \u00edndice e executar a mesma consulta utilizando um \u00edndice para comparar o desempenho dessa consulta nas duas situa\u00e7\u00f5es.<\/span><\/p>\n

Consulta sem utiliza\u00e7\u00e3o de \u00edndice<\/span><\/h3>\n

Para executar a consulta vamos criar uma tabela com 5GB e coletar as estat\u00edsticas da mesma.<\/span><\/p>\n

 <\/p>\n

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
<\/div><\/td>
SQL ><\/span> SELECT *<\/span> FROM V$VERSION<\/span> where rownum <<\/span> 2<\/span>;<\/span>
\n
\nBANNER
\n--------------------------------------------------------------------------------<\/span>
\nOracle Database 11g Enterprise Edition Release 11.<\/span>2.<\/span>0.<\/span>4.<\/span>0 -<\/span> 64bit Production
\n
\n1<\/span> row selected.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> alter session set workarea_size_policy=<\/span>MANUAL;<\/span>
\n
\nSession altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> alter session set sort_area_size=<\/span>2000000000<\/span>;<\/span>
\n
\nSession altered.<\/span>
\n
\nSQL ><\/span> create table dbtw01 as<\/span>
\n  2<\/span>     select 'DBTimeWizard - Performance and Tuning'<\/span> as<\/span> produto,<\/span>
\n  3<\/span>            mod(<\/span>rownum,<\/span>5<\/span>)<\/span>                            as<\/span> codigo,<\/span>
\n  4<\/span>            mod(<\/span>rownum,<\/span>1000<\/span>)<\/span>                         as<\/span> cliente_id ,<\/span>
\n  5<\/span>            mod(<\/span>rownum,<\/span>10000<\/span>)<\/span>                        as<\/span> postal_id,<\/span>
\n  6<\/span>            5000<\/span>                                     as<\/span> total_vendas,<\/span>
\n  7<\/span>            trunc(<\/span>sysdate -<\/span> 9999<\/span> +<\/span> mod(<\/span>rownum,<\/span>10000<\/span>)<\/span>)<\/span> as<\/span> data_venda
\n  8<\/span>       from dual connect by level<=<\/span>2e7<\/span>;<\/span>
\n
\nTable created.<\/span>
\n
\nSQL ><\/span> alter table dbtw01 nologging;<\/span>
\n
\nTable altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> insert \/*+ append *\/<\/span> into dbtw01 select *<\/span> from dbtw01;<\/span>
\n
\n20000000<\/span> rows created.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> commit;<\/span>
\n
\nCommit complete.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> insert \/*+ append *\/<\/span> into dbtw01 select *<\/span> from dbtw01;<\/span>
\n
\n40000000<\/span> rows created.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> commit;<\/span>
\n
\nCommit complete.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> select round<\/span><\/a>(<\/span>bytes\/<\/span>1024<\/span>\/<\/span>1024<\/span>\/<\/span>1024<\/span>,<\/span>1<\/span>)<\/span> as<\/span> "Tamanho GB"<\/span>
\n  2<\/span>    from user_segments
\n  3<\/span>   where segment_name =<\/span> 'DBTW01'<\/span>;<\/span>
\n
\nTamanho GB
\n----------<\/span>
\n       5.6<\/span>
\n
\n1<\/span> row selected.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> exec<\/span><\/a> dbms_stats.<\/span>gather_table_stats(<\/span>ownname=><\/span> USER,<\/span> tabname=><\/span> 'DBTW01'<\/span>,<\/span> cascade=><\/span> true<\/span>,<\/span> method_opt=><\/span> 'FOR ALL COLUMNS SIZE SKEWONLY'<\/span>)<\/span>;<\/span>
\n
\nPL\/<\/span>SQL procedure successfully completed.<\/span>
\n
\nSQL ><\/span> alter table dbtw01 logging;<\/span>
\n
\nTable altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> ALTER SESSION SET statistics_level=<\/span>ALL;<\/span>
\n
\nSession altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> set timing on;<\/span>
\nSQL ><\/span>
\nSQL ><\/span> select \/* NO_INDEX_DBTW01 *\/<\/span> min<\/span><\/a>(<\/span>cliente_id)<\/span> from dbtw01;<\/span>
\n
\nMIN<\/span><\/a>(<\/span>CLIENTE_ID)<\/span>
\n---------------<\/span>
\n              0<\/span>
\n
\n1<\/span> row selected.<\/span>
\n
\nElapsed:<\/span> 00<\/span>:<\/span>00<\/span>:<\/span>08.97<\/span>
\nSQL ><\/span>
\nSQL ><\/span> set timing off;<\/span>
\nSQL ><\/span>
\nSQL ><\/span> column sql_id new_value m_sql_id
\nSQL ><\/span> column child_number new_value m_child_no
\nSQL ><\/span>
\nSQL ><\/span> SELECT sql_id,<\/span> child_number
\n  2<\/span>   FROM v$sql<\/span>
\n  3<\/span>  WHERE sql_text LIKE '%NO_INDEX_DBTW01%'<\/span>
\n  4<\/span>    AND sql_text NOT LIKE '%v$sql%'<\/span>;<\/span>
\n
\nSQL_ID        CHILD_NUMBER
\n-------------<\/span> ------------<\/span>
\n2r831g60a8zx3            0<\/span>
\n
\n1<\/span> row selected.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span>
\nSQL ><\/span> SELECT *<\/span>
\n  2<\/span>   FROM TABLE (<\/span>dbms_xplan.<\/span>display_cursor (<\/span>'&m_sql_id'<\/span>,&<\/span>m_child_no,<\/span>'typical iostats last'<\/span>)<\/span>)<\/span>;<\/span>
\n
\nPLAN_TABLE_OUTPUT
\n----------------------------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\nSQL_ID  2r831g60a8zx3,<\/span> child number 0<\/span>
\n-------------------------------------<\/span>
\nselect \/* NO_INDEX_DBTW01 *\/<\/span> min<\/span><\/a>(<\/span>cliente_id)<\/span> from dbtw01
\n
\nPlan hash<\/span><\/a> value:<\/span> 2690647991<\/span>
\n
\n----------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n|<\/span> Id  |<\/span> Operation                  |<\/span> Name   |<\/span> Starts |<\/span> E-<\/span>Rows |<\/span>E-<\/span>Bytes|<\/span> Cost (<\/span>%<\/span>CPU)<\/span>|<\/span> E-<\/span>Time<\/span><\/a>   |<\/span> A-<\/span>Rows |<\/span>   A-<\/span>Time<\/span><\/a>   |<\/span> Buffers |<\/span> Reads  |<\/span>
\n----------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n|<\/span>   0<\/span> |<\/span> SELECT STATEMENT           |<\/span>        |<\/span>      1<\/span> |<\/span>        |<\/span>       |<\/span>   195K(<\/span>100<\/span>)<\/span>|<\/span>          |<\/span>      1<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>08.93<\/span> |<\/span>     719K|<\/span>    719K|<\/span>
\n|<\/span>   1<\/span> |<\/span>  SORT<\/span><\/a> AGGREGATE<\/span><\/a>            |<\/span>        |<\/span>      1<\/span> |<\/span>      1<\/span> |<\/span>     4<\/span> |<\/span>            |<\/span>          |<\/span>      1<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>08.93<\/span> |<\/span>     719K|<\/span>    719K|<\/span>
\n|<\/span>   2<\/span> |<\/span>   TABLE ACCESS STORAGE FULL|<\/span> DBTW01 |<\/span>      1<\/span> |<\/span>     80M|<\/span>   305M|<\/span>   195K  (<\/span>1<\/span>)<\/span>|<\/span> 00<\/span>:<\/span>39<\/span>:<\/span>09<\/span> |<\/span>     80M|<\/span>00<\/span>:<\/span>00<\/span>:<\/span>04.84<\/span> |<\/span>     719K|<\/span>    719K|<\/span>
\n----------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n
\n
\n14<\/span> rows selected.<\/span>
\n
\nSQL ><\/span><\/div><\/td><\/tr><\/tbody><\/table><\/div>\n

 <\/p>\n

No plano de execu\u00e7\u00e3o podemos observar que a consulta n\u00e3o utiliza \u00edndice de banco, ela faz um FULL TABLE SCAN no Exadata Storage e o tempo de execu\u00e7\u00e3o \u00e9 de aproximandamente 9 segundos.<\/span><\/p>\n

Consulta com utiliza\u00e7\u00e3o de \u00edndice<\/span><\/h3>\n

Agora vamos criar um \u00edndice na coluna “cliente_id” e executar a mesma consulta novamente:<\/span><\/p>\n

 <\/p>\n

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
<\/div><\/td>
SQL ><\/span> create index idx_dbtw01 on dbtw01(<\/span>cliente_id)<\/span>;<\/span>
\n
\nIndex created.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span>
\nSQL ><\/span> ALTER SESSION SET statistics_level=<\/span>ALL;<\/span>
\n
\nSession altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> set timing on;<\/span>
\nSQL ><\/span>
\nSQL ><\/span> select \/* WITH_INDEX_DBTW01 *\/<\/span> min<\/span><\/a>(<\/span>cliente_id)<\/span> from dbtw01;<\/span>
\n
\nMIN<\/span><\/a>(<\/span>CLIENTE_ID)<\/span>
\n---------------<\/span>
\n              0<\/span>
\n
\n1<\/span> row selected.<\/span>
\n
\nElapsed:<\/span> 00<\/span>:<\/span>00<\/span>:<\/span>00.01<\/span>
\nSQL ><\/span>
\nSQL ><\/span> set timing off;<\/span>
\nSQL ><\/span>
\nSQL ><\/span> column sql_id new_value m_sql_id
\nSQL ><\/span> column child_number new_value m_child_no
\nSQL ><\/span>
\nSQL ><\/span> SELECT sql_id,<\/span> child_number
\n  2<\/span>   FROM v$sql<\/span>
\n  3<\/span>  WHERE sql_text LIKE '%WITH_INDEX_DBTW01%'<\/span>
\n  4<\/span>    AND sql_text NOT LIKE '%v$sql%'<\/span>;<\/span>
\n
\nSQL_ID        CHILD_NUMBER
\n-------------<\/span> ------------<\/span>
\nb5wms71vkkmw1            0<\/span>
\n
\n1<\/span> row selected.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span>
\nSQL ><\/span> SELECT *<\/span>
\n  2<\/span>   FROM TABLE (<\/span>dbms_xplan.<\/span>display_cursor (<\/span>'&m_sql_id'<\/span>,&<\/span>m_child_no,<\/span>'typical iostats last'<\/span>)<\/span>)<\/span>;<\/span>
\n
\nPLAN_TABLE_OUTPUT
\n--------------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\nSQL_ID  b5wms71vkkmw1,<\/span> child number 0<\/span>
\n-------------------------------------<\/span>
\nselect \/* WITH_INDEX_DBTW01 *\/<\/span> min<\/span><\/a>(<\/span>cliente_id)<\/span> from dbtw01
\n
\nPlan hash<\/span><\/a> value:<\/span> 2619944286<\/span>
\n
\n--------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n|<\/span> Id  |<\/span> Operation                  |<\/span> Name       |<\/span> Starts |<\/span> E-<\/span>Rows |<\/span>E-<\/span>Bytes|<\/span> Cost (<\/span>%<\/span>CPU)<\/span>|<\/span> E-<\/span>Time<\/span><\/a>   |<\/span> A-<\/span>Rows |<\/span>   A-<\/span>Time<\/span><\/a>   |<\/span> Buffers |<\/span> Reads  |<\/span>
\n--------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n|<\/span>   0<\/span> |<\/span> SELECT STATEMENT           |<\/span>            |<\/span>      1<\/span> |<\/span>        |<\/span>       |<\/span>     3<\/span> (<\/span>100<\/span>)<\/span>|<\/span>          |<\/span>      1<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>00.01<\/span> |<\/span>       3<\/span> |<\/span>      2<\/span> |<\/span>
\n|<\/span>   1<\/span> |<\/span>  SORT<\/span><\/a> AGGREGATE<\/span><\/a>            |<\/span>            |<\/span>      1<\/span> |<\/span>      1<\/span> |<\/span>     4<\/span> |<\/span>            |<\/span>          |<\/span>      1<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>00.01<\/span> |<\/span>       3<\/span> |<\/span>      2<\/span> |<\/span>
\n|<\/span>   2<\/span> |<\/span>   INDEX FULL SCAN (<\/span>MIN<\/span><\/a>\/<\/span>MAX<\/span><\/a>)<\/span>|<\/span> IDX_DBTW01 |<\/span>      1<\/span> |<\/span>      1<\/span> |<\/span>     4<\/span> |<\/span>     3<\/span>   (<\/span>0<\/span>)<\/span>|<\/span> 00<\/span>:<\/span>00<\/span>:<\/span>01<\/span> |<\/span>      1<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>00.01<\/span> |<\/span>       3<\/span> |<\/span>      2<\/span> |<\/span>
\n--------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n
\n
\n14<\/span> rows selected.<\/span>
\n
\nSQL ><\/span><\/div><\/td><\/tr><\/tbody><\/table><\/div>\n

 <\/p>\n

No plano de execu\u00e7\u00e3o podemos observar que a consulta utiliza o \u00edndice criado e o tempo de resposta da consulta cai para menos de 1 cent\u00e9simo de segundo.<\/span><\/p>\n

Exadata Storage Indexes<\/span><\/h2>\n

O Exadata Storage Index (SI) \u00e9 um dos recursos que agiliza o acesso aos dados armazenados no storage, os segmentos armazenados no Exadata Storage s\u00e3o divididos em pequenas partes (default 1 MB) denominadas Storage regions e para cada uma delas s\u00e3o criados \u00edndices em memoria que armazenam os valores minimos e m\u00e1ximos para cada uma das colunas que s\u00e3o referenciadas na cla\u00fasula WHERE das instru\u00e7\u00f5es SQL. Esse indice \u00e9 utilizado para eliminar o acesso desnecess\u00e1rio ao disco, identificando as Storage regions que n\u00e3o possuem os valores que est\u00e3o sendo solicitados pelos filtros da cla\u00fasula WHERE.<\/span><\/p>\n

Porem a efic\u00e1cia do Storage Index e diretamente proporcional ao fator de clusteriza\u00e7\u00e3o das colunas nas Storage regions, ou seja, quanto mais organizados os dados estiverem nas colunas das Storage regions maior ser\u00e1 a efic\u00e1cia do SI para desprezar as regions que n\u00e3o precisam ser lidas.<\/span><\/p>\n

Desta forma existem colunas cujos valores e sequ\u00eancia de atualiza\u00e7\u00e3o n\u00e3o permitem a utiliza\u00e7\u00e3o eficaz do Storage Index, nestes casos a utiliza\u00e7\u00e3o do \u00edndice tradicional de banco de dados continua sendo uma \u00f3tima op\u00e7\u00e3o para melhoria da performance da instru\u00e7\u00e3o SQL.Veremos a seguir um exemplo pr\u00e1tico desta situa\u00e7\u00e3o.<\/span><\/p>\n

Consulta sem utiliza\u00e7\u00e3o de \u00edndice<\/span><\/h3>\n

Para executar a consulta vamos criar uma tabela com 5GB e coletar as estat\u00edsticas da mesma.<\/span><\/p>\n

 <\/p>\n

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
<\/div><\/td>
SQL ><\/span> SELECT *<\/span> FROM V$VERSION<\/span> where rownum <<\/span> 2<\/span>;<\/span>
\n
\nBANNER
\n--------------------------------------------------------------------------------<\/span>
\nOracle Database 11g Enterprise Edition Release 11.<\/span>2.<\/span>0.<\/span>4.<\/span>0 -<\/span> 64bit Production
\n
\n1<\/span> row selected.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> alter session set workarea_size_policy=<\/span>MANUAL;<\/span>
\n
\nSession altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> alter session set sort_area_size=<\/span>2000000000<\/span>;<\/span>
\n
\nSession altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> create table dbtw02 as<\/span>
\n  2<\/span>     select 'DBTimeWizard - Performance and Tuning'<\/span> as<\/span> produto,<\/span>
\n  3<\/span>            mod(<\/span>rownum,<\/span>5<\/span>)<\/span>                            as<\/span> codigo,<\/span>
\n  4<\/span>            mod(<\/span>rownum,<\/span>1000<\/span>)<\/span>                         as<\/span> cliente_id ,<\/span>
\n  5<\/span>            round<\/span><\/a>(<\/span>dbms_random.<\/span>value(<\/span>10000<\/span>,<\/span>0<\/span>)<\/span>)<\/span>        as<\/span> postal_id,<\/span>
\n  6<\/span>            5000<\/span>                                     as<\/span> total_vendas,<\/span>
\n  7<\/span>            trunc(<\/span>sysdate -<\/span> 9999<\/span> +<\/span> mod(<\/span>rownum,<\/span>10000<\/span>)<\/span>)<\/span> as<\/span> data_venda
\n  8<\/span>       from dual connect by level<=<\/span>2e7<\/span>;<\/span>
\n
\nTable created.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span>
\nSQL ><\/span> alter table dbtw02 nologging;<\/span>
\n
\nTable altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> insert \/*+ append *\/<\/span> into dbtw02 select *<\/span> from dbtw02;<\/span>
\n
\n20000000<\/span> rows created.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> commit;<\/span>
\n
\nCommit complete.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> insert \/*+ append *\/<\/span> into dbtw02 select *<\/span> from dbtw02;<\/span>
\n
\n40000000<\/span> rows created.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> commit;<\/span>
\n
\nCommit complete.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> select round<\/span><\/a>(<\/span>bytes\/<\/span>1024<\/span>\/<\/span>1024<\/span>\/<\/span>1024<\/span>,<\/span>1<\/span>)<\/span> as<\/span> "Tamanho GB"<\/span>
\n  2<\/span>    from user_segments
\n  3<\/span>   where segment_name =<\/span> 'DBTW02'<\/span>;<\/span>
\n
\nTamanho GB
\n----------<\/span>
\n       5.5<\/span>
\n
\n1<\/span> row selected.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> exec<\/span><\/a> dbms_stats.<\/span>gather_table_stats(<\/span>ownname=><\/span> USER,<\/span> tabname=><\/span> 'DBTW02'<\/span>,<\/span> cascade=><\/span> true<\/span>,<\/span> method_opt=><\/span> 'FOR ALL COLUMNS SIZE SKEWONLY'<\/span>)<\/span>;<\/span>
\n
\nPL\/<\/span>SQL procedure successfully completed.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> alter table dbtw02 logging;<\/span>
\n
\nTable altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> ALTER SESSION SET statistics_level=<\/span>ALL;<\/span>
\n
\nSession altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> set timing on;<\/span>
\nSQL ><\/span>
\nSQL ><\/span> select \/* NO_INDEX_DBTW02 *\/<\/span> *<\/span>
\n  2<\/span>    from dbtw02
\n  3<\/span>   where cliente_id =<\/span> 711<\/span>
\n  4<\/span>     and postal_id =<\/span> 7348<\/span>;<\/span>
\n
\nPRODUTO                                   CODIGO CLIENTE_ID  POSTAL_ID TOTAL_VENDAS DATA_VEND
\n-------------------------------------<\/span> ----------<\/span> ----------<\/span> ----------<\/span> ------------<\/span> ---------<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 02<\/span>-<\/span>JAN-<\/span>13<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 02<\/span>-<\/span>JAN-<\/span>13<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 02<\/span>-<\/span>JAN-<\/span>13<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 02<\/span>-<\/span>JAN-<\/span>13<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\n
\n16<\/span> rows selected.<\/span>
\n
\nElapsed:<\/span> 00<\/span>:<\/span>00<\/span>:<\/span>22.22<\/span>
\nSQL ><\/span>
\nSQL ><\/span> set timing off;<\/span>
\nSQL ><\/span>
\nSQL ><\/span> column sql_id new_value m_sql_id
\nSQL ><\/span> column child_number new_value m_child_no
\nSQL ><\/span>
\nSQL ><\/span> SELECT sql_id,<\/span> child_number
\n  2<\/span>   FROM v$sql<\/span>
\n  3<\/span>  WHERE sql_text LIKE '%NO_INDEX_DBTW02%'<\/span>
\n  4<\/span>    AND sql_text NOT LIKE '%v$sql%'<\/span>;<\/span>
\n
\nSQL_ID        CHILD_NUMBER
\n-------------<\/span> ------------<\/span>
\n6hs5v87115nss            0<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span>
\nSQL ><\/span> SELECT *<\/span>
\n  2<\/span>   FROM TABLE (<\/span>dbms_xplan.<\/span>display_cursor (<\/span>'&m_sql_id'<\/span>,&<\/span>m_child_no,<\/span>'typical iostats last'<\/span>)<\/span>)<\/span>;<\/span>
\nold   2<\/span>:<\/span>  FROM TABLE (<\/span>dbms_xplan.<\/span>display_cursor (<\/span>'&m_sql_id'<\/span>,&<\/span>m_child_no,<\/span>'typical iostats last'<\/span>)<\/span>)<\/span>
\nnew<\/span>   2<\/span>:<\/span>  FROM TABLE (<\/span>dbms_xplan.<\/span>display_cursor (<\/span>'6hs5v87115nss'<\/span>,<\/span>         0<\/span>,<\/span>'typical iostats last'<\/span>)<\/span>)<\/span>
\n
\nPLAN_TABLE_OUTPUT
\n-----------------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\nSQL_ID  6hs5v87115nss,<\/span> child number 0<\/span>
\n-------------------------------------<\/span>
\nselect \/* NO_INDEX_DBTW02 *\/<\/span> *<\/span>   from dbtw02  where cliente_id =<\/span> 711<\/span>
\nand postal_id =<\/span> 7348<\/span>
\n
\nPlan hash<\/span><\/a> value:<\/span> 202550308<\/span>
\n
\n---------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n|<\/span> Id  |<\/span> Operation                 |<\/span> Name   |<\/span> Starts |<\/span> E-<\/span>Rows |<\/span>E-<\/span>Bytes|<\/span> Cost (<\/span>%<\/span>CPU)<\/span>|<\/span> E-<\/span>Time<\/span><\/a>   |<\/span> A-<\/span>Rows |<\/span>   A-<\/span>Time<\/span><\/a>   |<\/span> Buffers |<\/span> Reads  |<\/span>
\n---------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n|<\/span>   0<\/span> |<\/span> SELECT STATEMENT          |<\/span>        |<\/span>      1<\/span> |<\/span>        |<\/span>       |<\/span>   195K(<\/span>100<\/span>)<\/span>|<\/span>          |<\/span>     16<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>22.20<\/span> |<\/span>     719K|<\/span>    455K|<\/span>
\n|*<\/span>  1<\/span> |<\/span>  TABLE ACCESS STORAGE FULL|<\/span> DBTW02 |<\/span>      1<\/span> |<\/span>      8<\/span> |<\/span>   480<\/span> |<\/span>   195K  (<\/span>1<\/span>)<\/span>|<\/span> 00<\/span>:<\/span>39<\/span>:<\/span>10<\/span> |<\/span>     16<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>22.20<\/span> |<\/span>     719K|<\/span>    455K|<\/span>
\n---------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n
\nPredicate Information (<\/span>identified by operation id)<\/span>:<\/span>
\n---------------------------------------------------<\/span>
\n
\n   1<\/span> -<\/span> storage(<\/span>(<\/span>"POSTAL_ID"<\/span>=<\/span>7348<\/span> AND "CLIENTE_ID"<\/span>=<\/span>711<\/span>)<\/span>)<\/span>
\n       filter(<\/span>(<\/span>"POSTAL_ID"<\/span>=<\/span>7348<\/span> AND "CLIENTE_ID"<\/span>=<\/span>711<\/span>)<\/span>)<\/span>
\n
\n
\n20<\/span> rows selected.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span>
\nSQL ><\/span> select name,<\/span> value
\n  2<\/span>    from v$mystat<\/span> s,<\/span> v$statname<\/span> n
\n  3<\/span>   where s.<\/span>statistic# = n.statistic#
\n<\/span>  4<\/span>     and name like '%storage index%'<\/span>;<\/span>
\n
\nNAME                                                                  VALUE
\n----------------------------------------------------------------<\/span> ----------<\/span>
\ncell physical IO bytes saved by storage index                             0<\/span>
\n
\nSQL ><\/span><\/div><\/td><\/tr><\/tbody><\/table><\/div>\n

 <\/p>\n

Examinando o plano de execu\u00e7\u00e3o acima verificamos que a consulta fez um FTS e retornou o resultado ap\u00f3s 22 segundos, consultando as vis\u00f5es de estat\u00edticas do banco podemos observar tambem que o Storage Index n\u00e3o foi utilizado.<\/span><\/p>\n

Consulta com utiliza\u00e7\u00e3o de \u00edndice<\/span><\/h3>\n

Agora vamos criar um \u00edndice na coluna “postal_id” e executar a mesma consulta novamente:<\/span><\/p>\n

 <\/p>\n

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
<\/div><\/td>
SQL ><\/span> create index idx_dbtw02 on dbtw02(<\/span>postal_id)<\/span> parallel 16<\/span>;<\/span>
\n
\nIndex created.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> alter index idx_dbtw02 noparallel;<\/span>
\n
\nIndex altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> ALTER SESSION SET statistics_level=<\/span>ALL;<\/span>
\n
\nSession altered.<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span> set timing on;<\/span>
\nSQL ><\/span>
\nSQL ><\/span> select \/* WITH_INDEX_DBTW02 *\/<\/span> *<\/span>
\n  2<\/span>    from dbtw02
\n  3<\/span>   where cliente_id =<\/span> 711<\/span>
\n  4<\/span>     and postal_id =<\/span> 7348<\/span>;<\/span>
\n
\nPRODUTO                                   CODIGO CLIENTE_ID  POSTAL_ID TOTAL_VENDAS DATA_VEND
\n-------------------------------------<\/span> ----------<\/span> ----------<\/span> ----------<\/span> ------------<\/span> ---------<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 02<\/span>-<\/span>JAN-<\/span>13<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 02<\/span>-<\/span>JAN-<\/span>13<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 02<\/span>-<\/span>JAN-<\/span>13<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 02<\/span>-<\/span>JAN-<\/span>13<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\nDBTimeWizard -<\/span> Performance and Tuning          1<\/span>        711<\/span>       7348<\/span>         5000<\/span> 29<\/span>-<\/span>SEP-<\/span>15<\/span>
\n
\n16<\/span> rows selected.<\/span>
\n
\nElapsed:<\/span> 00<\/span>:<\/span>00<\/span>:<\/span>00.18<\/span>
\nSQL ><\/span>
\nSQL ><\/span>
\nSQL ><\/span> set timing off;<\/span>
\nSQL ><\/span>
\nSQL ><\/span> column sql_id new_value m_sql_id
\nSQL ><\/span> column child_number new_value m_child_no
\nSQL ><\/span>
\nSQL ><\/span> SELECT sql_id,<\/span> child_number
\n  2<\/span>   FROM v$sql<\/span>
\n  3<\/span>  WHERE sql_text LIKE '%WITH_INDEX_DBTW02%'<\/span>
\n  4<\/span>    AND sql_text NOT LIKE '%v$sql%'<\/span>;<\/span>
\n
\nSQL_ID        CHILD_NUMBER
\n-------------<\/span> ------------<\/span>
\n901fc75zzrnrq            0<\/span>
\n
\nSQL ><\/span>
\nSQL ><\/span>
\nSQL ><\/span> SELECT *<\/span>
\n  2<\/span>   FROM TABLE (<\/span>dbms_xplan.<\/span>display_cursor (<\/span>'&m_sql_id'<\/span>,&<\/span>m_child_no,<\/span>'typical iostats last'<\/span>)<\/span>)<\/span>;<\/span>
\nold   2<\/span>:<\/span>  FROM TABLE (<\/span>dbms_xplan.<\/span>display_cursor (<\/span>'&m_sql_id'<\/span>,&<\/span>m_child_no,<\/span>'typical iostats last'<\/span>)<\/span>)<\/span>
\nnew<\/span>   2<\/span>:<\/span>  FROM TABLE (<\/span>dbms_xplan.<\/span>display_cursor (<\/span>'901fc75zzrnrq'<\/span>,<\/span>         0<\/span>,<\/span>'typical iostats last'<\/span>)<\/span>)<\/span>
\n
\nPLAN_TABLE_OUTPUT
\n---------------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\nSQL_ID  901fc75zzrnrq,<\/span> child number 0<\/span>
\n-------------------------------------<\/span>
\nselect \/* WITH_INDEX_DBTW02 *\/<\/span> *<\/span>   from dbtw02  where cliente_id =<\/span> 711<\/span>
\n  and postal_id =<\/span> 7348<\/span>
\n
\nPlan hash<\/span><\/a> value:<\/span> 388684596<\/span>
\n
\n---------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n|<\/span> Id  |<\/span> Operation                   |<\/span> Name       |<\/span> Starts |<\/span> E-<\/span>Rows |<\/span>E-<\/span>Bytes|<\/span> Cost (<\/span>%<\/span>CPU)<\/span>|<\/span> E-<\/span>Time<\/span><\/a>   |<\/span> A-<\/span>Rows |<\/span>   A-<\/span>Time<\/span><\/a>   |<\/span> Buffers |<\/span> Reads  |<\/span>
\n---------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n|<\/span>   0<\/span> |<\/span> SELECT STATEMENT            |<\/span>            |<\/span>      1<\/span> |<\/span>        |<\/span>       |<\/span>  7977<\/span> (<\/span>100<\/span>)<\/span>|<\/span>          |<\/span>     16<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>00.17<\/span> |<\/span>    7957<\/span> |<\/span>    641<\/span> |<\/span>
\n|*<\/span>  1<\/span> |<\/span>  TABLE ACCESS BY INDEX ROWID|<\/span> DBTW02     |<\/span>      1<\/span> |<\/span>      8<\/span> |<\/span>   480<\/span> |<\/span>  7977<\/span>   (<\/span>1<\/span>)<\/span>|<\/span> 00<\/span>:<\/span>01<\/span>:<\/span>36<\/span> |<\/span>     16<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>00.17<\/span> |<\/span>    7957<\/span> |<\/span>    641<\/span> |<\/span>
\n|*<\/span>  2<\/span> |<\/span>   INDEX RANGE<\/span><\/a> SCAN          |<\/span> IDX_DBTW02 |<\/span>      1<\/span> |<\/span>   7999<\/span> |<\/span>       |<\/span>    19<\/span>   (<\/span>0<\/span>)<\/span>|<\/span> 00<\/span>:<\/span>00<\/span>:<\/span>01<\/span> |<\/span>   7984<\/span> |<\/span>00<\/span>:<\/span>00<\/span>:<\/span>00.01<\/span> |<\/span>      22<\/span> |<\/span>     19<\/span> |<\/span>
\n---------------------------------------------------------------------------------------------------------------------------------------------<\/span>
\n
\nPredicate Information (<\/span>identified by operation id)<\/span>:<\/span>
\n---------------------------------------------------<\/span>
\n
\n   1<\/span> -<\/span> filter(<\/span>"CLIENTE_ID"<\/span>=<\/span>711<\/span>)<\/span>
\n   2<\/span> -<\/span> access(<\/span>"POSTAL_ID"<\/span>=<\/span>7348<\/span>)<\/span>
\n
\n
\n21<\/span> rows selected.<\/span>
\n
\nSQL ><\/span><\/div><\/td><\/tr><\/tbody><\/table><\/div>\n

 <\/p>\n

No plano de execu\u00e7\u00e3o acima constatamos que a consulta utilizou o \u00edndice criado e terminou em apenas 17 cent\u00e9simo de segundos.\u00a0<\/span><\/p>\n

Conclus\u00e3o<\/span><\/h2>\n

Os dois exemplos acima mostram que eliminar \u00edndices de forma indiscriminada ap\u00f3s uma migra\u00e7\u00e3o para o Exadata n\u00e3o \u00e9 uma estrat\u00e9gia adequada em termos de performance, mas tambem encontramos situa\u00e7\u00f5es no ambiente Exadata em que a n\u00e3o utiliza\u00e7\u00e3o do \u00edndice melhora a performance da instru\u00e7\u00e3o SQL, portanto a estrat\u00e9gia mais adequada \u00e9 analisar os casos onde o tempo de resposta \u00e9 insatisfat\u00f3rio e fazer o ajustes necess\u00e1rios.<\/span><\/p>\n

Refer\u00eancias<\/span><\/h2>\n

https:\/\/richardfoote.wordpress.com\/2012\/12\/18\/storage-indexes-vs-database-indexes-part-i-minmax-maxwells-silver-hammer\/<\/a><\/span><\/strong><\/p>\n

https:\/\/richardfoote.wordpress.com\/2012\/12\/19\/storage-indexes-vs-database-indexes-part-ii-clustering-factor-fast-track\/<\/a><\/span><\/strong><\/p>\n

http:\/\/www.oaktable.net\/blog\/when-exadata%E2%80%99s-storage-indexes-used?page=2<\/a><\/span><\/strong>
\n 
\n 
\n\"Promo-D75L\"<\/a><\/p>\n

\"video-02\"<\/a>
\n 
\n 
\n 
\n 
\n 
\n 
\n 
\n 
\n 
\n 
\n 
\n 
\n 
\n <\/p>\n","protected":false},"excerpt":{"rendered":"

\u00c9 comum ouvir outras pessoas dizendo que no Exadata n\u00e3o precisamos mais dos \u00edndices de banco de dados tradicionais pois o Exadata responde melhor sem eles e que podemos excluir todos os \u00edndices para liberar espa\u00e7o no Storage. Na pr\u00e1tica as coisas s\u00e3o um pouco diferentes, existem muitos cen\u00e1rios onde os \u00edndices de banco de dados ainda s\u00e3o cr\u00edticos para<\/p>\n","protected":false},"author":2,"featured_media":810,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[142],"tags":[146,151,143,145,144,149],"class_list":["post-809","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-exadata","tag-cell_offload_processing","tag-database-index","tag-exadata","tag-smart-scan","tag-storage-cell","tag-storage-index"],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/posts\/809","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/comments?post=809"}],"version-history":[{"count":16,"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/posts\/809\/revisions"}],"predecessor-version":[{"id":2092,"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/posts\/809\/revisions\/2092"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/media\/810"}],"wp:attachment":[{"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/media?parent=809"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/categories?post=809"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dbtimewizard.com.br\/blog\/wp-json\/wp\/v2\/tags?post=809"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}