关于分区索引与全局索引性能比较的示例
2012-12-31 13:47
225 查看
说明：之前使用range分区做出来的效果不明显，这次使用hash分区。
1、准备工作：
2、分区索引性能优于全局索引的例子：
3、分区索引性能低于全局索引的例子1：
分区索引性能低于全局索引的例子2：
1、准备工作：
----创建两张一样的hash分区表，jacks_part和echos_part------------------ 1 SQL> create table jacks_part (owner varchar2(30),object_id number,object_name varchar2(128)) 2 partition by hash(object_id) 3 partitions 30; Table created. SQL> create table echos_part (owner varchar2(30),object_id number,object_name varchar2(128)) 2 partition by hash(object_id) 3 partitions 30; Table created. ----分别向两张表插入一些记录----------------- 13 SQL> insert into jacks_part select owner,object_id,object_name from dba_objects; 72196 rows created. SQL> insert into echos_part select owner,object_id,object_name from jacks_part; 72196 rows created. SQL> commit; Commit complete. ----分别创建global索引和local索引--------------- 25 SQL> create index globals_ind on jacks_part(object_id) 2 global partition by hash(object_id); Index created. SQL> create index locals_ind on echos_part(object_id) local; Index created. ----查询索引是否正确-------------------------- 34 SQL> select index_name,table_name,locality from user_part_indexes; INDEX_NAME TABLE_NAME LOCALI ------------------ ------------------------------ ------ LOCALS_IND ECHOS_PART LOCAL GLOBALS_IND JACKS_PART GLOBAL
2、分区索引性能优于全局索引的例子：
SQL> set linesize 200;
SQL> set autotrace traceonly;
SQL> select /*+ index(echos_part,locals_ind) */ * from echos_part where object_id>100;
72097 rows selected.
Execution Plan
----------------------------------------------------------
Plan hash value: 3092815211
-----------------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | Pstart| Pstop |
-----------------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 4228 | 396K| 89 (0)| 00:00:02 | | |
| 1 | PARTITION HASH ALL | | 4228 | 396K| 89 (0)| 00:00:02 | 1 | 30 |
| 2 | TABLE ACCESS BY LOCAL INDEX ROWID| ECHOS_PART | 4228 | 396K| 89 (0)| 00:00:02 | 1 | 30 |
|* 3 | INDEX RANGE SCAN | LOCALS_IND | 4228 | | 25 (0)| 00:00:01 | 1 | 30 |
-----------------------------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
3 - access("OBJECT_ID">100)
Note
-----
- dynamic sampling used for this statement (level=2)
Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
10562 consistent gets
0 physical reads
0 redo size
3128267 bytes sent via SQL*Net to client
53285 bytes received via SQL*Net from client
4808 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
72097 rows processed
SQL> select /*+ index(jacks_part,globals_ind) */ * from jacks_part where object_id>100;
72097 rows selected.
Execution Plan
----------------------------------------------------------
Plan hash value: 2501448352
-------------------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | Pstart| Pstop |
-------------------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 2500 | 234K| 4639 (1)| 00:00:56 | | |
| 1 | PARTITION HASH SINGLE | | 2500 | 234K| 4639 (1)| 00:00:56 | 1 | 1 |
| 2 | TABLE ACCESS BY GLOBAL INDEX ROWID| JACKS_PART | 2500 | 234K| 4639 (1)| 00:00:56 | ROWID | ROWID |
|* 3 | INDEX RANGE SCAN | GLOBALS_IND | 2500 | | 15 (0)| 00:00:01 | 1 | 1 |
-------------------------------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
3 - access("OBJECT_ID">100)
Note
-----
- dynamic sampling used for this statement (level=2)
Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
74718 consistent gets
0 physical reads
0 redo size
3077218 bytes sent via SQL*Net to client
53285 bytes received via SQL*Net from client
4808 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
72097 rows processed3、分区索引性能低于全局索引的例子1：
SQL> select /*+ index(echos_part,locals_ind) */ count(*) from echos_part where object_id>100;
Execution Plan
----------------------------------------------------------
Plan hash value: 2317569636
--------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | Pstart| Pstop |
--------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 13 | 25 (0)| 00:00:01 | | |
| 1 | SORT AGGREGATE | | 1 | 13 | | | | |
| 2 | PARTITION HASH ALL| | 4228 | 54964 | 25 (0)| 00:00:01 | 1 | 30 |
|* 3 | INDEX RANGE SCAN | LOCALS_IND | 4228 | 54964 | 25 (0)| 00:00:01 | 1 | 30 |
--------------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
3 - access("OBJECT_ID">100)
Note
-----
- dynamic sampling used for this statement (level=2)
Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
205 consistent gets
0 physical reads
0 redo size
424 bytes sent via SQL*Net to client
419 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed
SQL> select /*+ index(jacks_part,globals_ind) */ count(*) from jacks_part where object_id>100;
Execution Plan
----------------------------------------------------------
Plan hash value: 2478129137
------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | Pstart| Pstop |
------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 13 | 15 (0)| 00:00:01 | | |
| 1 | SORT AGGREGATE | | 1 | 13 | | | | |
| 2 | PARTITION HASH SINGLE| | 2500 | 32500 | 15 (0)| 00:00:01 | 1 | 1 |
|* 3 | INDEX RANGE SCAN | GLOBALS_IND | 2500 | 32500 | 15 (0)| 00:00:01 | 1 | 1 |
------------------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
3 - access("OBJECT_ID">100)
Note
-----
- dynamic sampling used for this statement (level=2)
Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
201 consistent gets
0 physical reads
0 redo size
424 bytes sent via SQL*Net to client
419 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed分区索引性能低于全局索引的例子2：
SQL> drop index globals_ind;
Index dropped.
SQL> create index global_indexs on jacks_part(object_id) global;
Index created.
SQL> select /*+ index(echos_part,locals_ind) */ count(*) from echos_part where object_id>100;
Execution Plan
----------------------------------------------------------
Plan hash value: 2317569636
--------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | Pstart| Pstop |
--------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 5 | 175 (0)| 00:00:03 | | |
| 1 | SORT AGGREGATE | | 1 | 5 | | | | |
| 2 | PARTITION HASH ALL| | 72101 | 352K| 175 (0)| 00:00:03 | 1 | 30 |
|* 3 | INDEX RANGE SCAN | LOCALS_IND | 72101 | 352K| 175 (0)| 00:00:03 | 1 | 30 |
--------------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
3 - access("OBJECT_ID">100)
Statistics
----------------------------------------------------------
1704 recursive calls
0 db block gets
437 consistent gets
206 physical reads
0 redo size
SQL> select /*+ index(jacks_part,global_indexs) */ count(*) from jacks_part where object_id>100;
Execution Plan
----------------------------------------------------------
Plan hash value: 1016566238
-----------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
-----------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 5 | 201 (0)| 00:00:03 |
| 1 | SORT AGGREGATE | | 1 | 5 | | |
|* 2 | INDEX RANGE SCAN| GLOBAL_INDEXS | 72101 | 352K| 201 (0)| 00:00:03 |
-----------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
2 - access("OBJECT_ID">100)
Statistics
----------------------------------------------------------
1 recursive calls
0 db block gets
201 consistent gets
200 physical reads
0 redo size
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
- 分区索引--本地索引和全局索引比较
- Oracle分区索引--本地索引和全局索引比较
- Oracle分区索引--本地索引和全局索引比较
- 分区索引--本地索引和全局索引比较
- Oracle分区索引-本地索引和全局索引比较
- 分区索引--本地索引和全局索引比较
- 分区索引--本地索引和全局索引比较
- 分区索引--本地索引和全局索引比较
- 全局分区索引和本地索引示例
- Oracle分区索引--本地索引和全局索引比较
- Oracle12c中性能优化&功能增强新特性之全局索引DROP和TRUNCATE 分区的异步维护
- Oracle 12C 新特性之表分区带异步全局索引异步维护（一次add、truncate、drop、spilt、merge多个分区）
- Oracle B-tree、位图、全文索引三大索引性能比较及优缺点汇总
- 全局索引与本地分区索引的一个重要区别
- Oracle 12C 新特性之表分区带 异步全局索引异步维护（一次add、truncate、drop、spilt、merge多个分区）
- 全局分区索引与局部分区索引
- 关于Polar Code和高阶调制16QAM与Turbｏ与高阶调制性能的比较
- 关于LinkedList和ArrayList 插入元素的性能比较疑惑，知道的请指教......
- 关于count(*)得 非聚合索引 性能大于聚合索引性能研究