Home  >  Article  >  Database  >  数据库设计--数据的垂直拆分

数据库设计--数据的垂直拆分

WBOY
WBOYOriginal
2016-06-07 15:56:511244browse

如果表字段太多,如果表中有些字段比较大,即便是你只查有限的几个字段,在做表关联和全表扫的时候,由于扫描的数据块多,性能方面还是会不理想。因为oracle扫描的时候是按照块为单位扫描,读取的时候也是按块为单位读取,所以这种功能无法在SQL层面上优化的

如果表字段太多,如果表中有些字段比较大,即便是你只查有限的几个字段,在做表关联和全表扫的时候,由于扫描的数据块多,性能方面还是会不理想。因为oracle扫描的时候是按照块为单位扫描,读取的时候也是按块为单位读取,所以这种功能无法在SQL层面上优化的时候,可以考虑做数据的垂直切分,下面来做个试验:

--制造数据不做垂直切分
create table test(
a number,
b varchar2(4000),
c varchar2(4000),
d varchar2(4000),
e varchar2(4000),
f varchar2(4000),
g varchar2(4000),
h varchar2(4000)
);
INSERT INTO test
SELECT ROWNUM,
rpad('*', 4000, 1),
rpad('*', 4000, 1),
rpad('*', 4000, 1),
rpad('*', 4000, 1),
rpad('*', 4000, 1),
rpad('*', 4000, 1),
rpad('*', 4000, 1)
FROM DUAL
CONNECT BY ROWNUM commit;
create table test1 as select * from test;

--制造数据做垂直切分
create table test_cuizhi(
a number
);
INSERT INTO test_cuizhi
SELECT ROWNUM
FROM DUAL
CONNECT BY ROWNUM commit;
create table test_cuizhi1 as select * from test_cuizhi;

--开始测试,只是取两个最小的字段
SQL> set timing on
SQL> set autotrace traceonly
SQL> select t.a,t1.a from test t, test1 t1 where t.a=t1.a;
已选择100000行。
已用时间: 00: 00: 53.17
执行计划
----------------------------------------------------------
Plan hash value: 2400077556
----------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
----------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 44504 | 1129K| 173K (1)| 00:34:38 |
|* 1 | HASH JOIN | | 44504 | 1129K| 173K (1)| 00:34:38 |
| 2 | TABLE ACCESS FULL| TEST | 44504 | 564K| 87801 (1)| 00:17:34 |
| 3 | TABLE ACCESS FULL| TEST1 | 117K| 1490K| 85344 (1)| 00:17:05 |
----------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - access("T"."A"="T1"."A")
Note
-----
- dynamic sampling used for this statement
统计信息
----------------------------------------------------------
52 recursive calls
0 db block gets
795627 consistent gets
534917 physical reads
0 redo size
1664840 bytes sent via SQL*Net to client
73664 bytes received via SQL*Net from client
6668 SQL*Net roundtrips to/from client
2 sorts (memory)
0 sorts (disk)
100000 rows processed
SQL> /
已选择100000行。
已用时间: 00: 00: 33.36
执行计划
----------------------------------------------------------
Plan hash value: 2400077556
----------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
----------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 44504 | 1129K| 173K (1)| 00:34:38 |
|* 1 | HASH JOIN | | 44504 | 1129K| 173K (1)| 00:34:38 |
| 2 | TABLE ACCESS FULL| TEST | 44504 | 564K| 87801 (1)| 00:17:34 |
| 3 | TABLE ACCESS FULL| TEST1 | 117K| 1490K| 85344 (1)| 00:17:05 |
----------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - access("T"."A"="T1"."A")
Note
-----
- dynamic sampling used for this statement
统计信息
----------------------------------------------------------
0 recursive calls
0 db block gets
795446 consistent gets
552087 physical reads
0 redo size
1664840 bytes sent via SQL*Net to client
73664 bytes received via SQL*Net from client
6668 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
100000 rows processed


SQL> select t.a,t1.a from test_cuizhi t, test_cuizhi1 t1 where t.a=t1.a;
已选择100000行。
已用时间: 00: 00: 06.17
执行计划
----------------------------------------------------------
Plan hash value: 2501302817
-------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)| Time |
-------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 88629 | 2250K| | 310 (2)| 00:00:04 |
|* 1 | HASH JOIN | | 88629 | 2250K| 2168K| 310 (2)| 00:00:04 |
| 2 | TABLE ACCESS FULL| TEST_CUIZHI | 88629 | 1125K| | 42 (3)| 00:00:01 |
| 3 | TABLE ACCESS FULL| TEST_CUIZHI1 | 101K| 1288K| | 39 (3)| 00:00:01 |
-------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - access("T"."A"="T1"."A")
Note
-----
- dynamic sampling used for this statement


统计信息
----------------------------------------------------------
52 recursive calls
0 db block gets
7139 consistent gets
153 physical reads
0 redo size
1664840 bytes sent via SQL*Net to client
73664 bytes received via SQL*Net from client
6668 SQL*Net roundtrips to/from client
2 sorts (memory)
0 sorts (disk)
100000 rows processed


SQL> /
已选择100000行。
已用时间: 00: 00: 06.06
执行计划
----------------------------------------------------------
Plan hash value: 2501302817
-------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)| Time |
-------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 88629 | 2250K| | 310 (2)| 00:00:04 |
|* 1 | HASH JOIN | | 88629 | 2250K| 2168K| 310 (2)| 00:00:04 |
| 2 | TABLE ACCESS FULL| TEST_CUIZHI | 88629 | 1125K| | 42 (3)| 00:00:01 |
| 3 | TABLE ACCESS FULL| TEST_CUIZHI1 | 101K| 1288K| | 39 (3)| 00:00:01 |
-------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - access("T"."A"="T1"."A")
Note
-----
- dynamic sampling used for this statement
统计信息
----------------------------------------------------------
0 recursive calls
0 db block gets
7008 consistent gets
0 physical reads
0 redo size
1664840 bytes sent via SQL*Net to client
73664 bytes received via SQL*Net from client
6668 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
100000 rows processed
Statement:
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn