oracle index logging,index在logging什么？

index在logging什么？

比较疑惑，create index 时默认设置logging,不明白索引需要logging些什么？

我想只要表logging了，索引有定义，就可以恢复么，干吗还要再logging呢？

如果index是logging的，那它的log又是怎么记录的呢？

更奇怪下面的一次测试：

SQL> create table test1 as select * from dba_objects where 1<>1;

Table created

SQL> create index INX_1 on test1 (object_id);

Index created

SQL> insert into test1 select * from dba_objects;

已创建30811行。

Execution Plan

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

0      INSERT STATEMENT Optimizer=CHOOSE

1    0   VIEW OF 'DBA_OBJECTS'

2    1     UNION-ALL

3    2       FILTER

4    3         TABLE ACCESS (BY INDEX ROWID) OF 'OBJ$'

5    4           NESTED LOOPS

6    5             TABLE ACCESS (FULL) OF 'USER$'

7    5             INDEX (RANGE SCAN) OF 'I_OBJ2' (UNIQUE)

8    3         TABLE ACCESS (BY INDEX ROWID) OF 'IND$'

9    8           INDEX (UNIQUE SCAN) OF 'I_IND1' (UNIQUE)

10    2       TABLE ACCESS (BY INDEX ROWID) OF 'LINK$'

11   10         NESTED LOOPS

12   11           TABLE ACCESS (FULL) OF 'USER$'

13   11           INDEX (RANGE SCAN) OF 'I_LINK1' (NON-UNIQUE)

Statistics

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

616  recursive calls

37587  db block gets

25032  consistent gets

5  physical reads

8635516  redo size

622  bytes sent via SQL*Net to client

543  bytes received via SQL*Net from client

3  SQL*Net roundtrips to/from client

1  sorts (memory)

0  sorts (disk)

30811  rows processed

SQL> truncate table test1;

表已截掉。

SQL> alter index x_id nologging;

索引已更改。

SQL> insert into test1 select * from dba_objects;

已创建30811行。

Execution Plan

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

0      INSERT STATEMENT Optimizer=CHOOSE

1    0   VIEW OF 'DBA_OBJECTS'

2    1     UNION-ALL

3    2       FILTER

4    3         TABLE ACCESS (BY INDEX ROWID) OF 'OBJ$'

5    4           NESTED LOOPS

6    5             TABLE ACCESS (FULL) OF 'USER$'

7    5             INDEX (RANGE SCAN) OF 'I_OBJ2' (UNIQUE)

8    3         TABLE ACCESS (BY INDEX ROWID) OF 'IND$'

9    8           INDEX (UNIQUE SCAN) OF 'I_IND1' (UNIQUE)

10    2       TABLE ACCESS (BY INDEX ROWID) OF 'LINK$'

11   10         NESTED LOOPS

12   11           TABLE ACCESS (FULL) OF 'USER$'

13   11           INDEX (RANGE SCAN) OF 'I_LINK1' (NON-UNIQUE)

Statistics

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

817  recursive calls

37634  db block gets

25068  consistent gets

5  physical reads

8654080  redo size

622  bytes sent via SQL*Net to client

543  bytes received via SQL*Net from client

3  SQL*Net roundtrips to/from client

8  sorts (memory)

0  sorts (disk)

30811  rows processed

怎么index在nologging时的redo(logging:8635516 ; nologging:8654080)反而更多呢？

请兄弟们指教，谢谢。