感受ORACLE 12C IN_MEMORY性能威力
oracle 12C in_memory(简称IM)对数据库性能的提升是巨大的，最重要的一点是这一切
对应用程序是透明的。首先我们来看看实验中IM的相关参数设置
使用imemory特性的前提，要设置这两个参数
SQL>alter system set inmemory_max_populate_servers=2 scope=spfile;
SQL>alter system set inmemory_size=100M scope=spfile;(这个参数设置的大小最小为100M)
SQL> show parameter inmemory;
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
inmemory_clause_default string
inmemory_force string DEFAULT
inmemory_max_populate_servers integer 2
inmemory_query string ENABLE
inmemory_size big integer 100M
inmemory_trickle_repopulate_servers_ integer 1
percent
optimizer_inmemory_aware boolean TRUE
创建一个普通用户，创建一个表数据来自dba_objects;
SQL> create user c##frank identified by frank;
User created.
SQL> grant connect,resource,unlimited tablespace to c##frank;
Grant succeeded.
SQL> create table c##frank.t1 as select * from dba_objects;
Table created.
把用户表放到IM中.
SQL> alter table t1 c##frank.inmemory;
Table altered.
收集一下统计息.
SQL> exec dbms_stats.gather_table_stats('C##FRANK','T1');
PL/SQL procedure successfully completed.
SQL> set autotrace on;
SQL> select count(object_id) from c##frank.t1;
COUNT(OBJECT_ID)
----------------
90905
Execution Plan
----------------------------------------------------------
Plan hash value: 3724264953
------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 5 | 27 (0)| 00:00:01 |
| 1 | SORT AGGREGATE | | 1 | 5 | | |
| 2 | TABLE ACCESS INMEMORY FULL| T1 | 90905 | 443K| 27 (0)| 00:00:01 |
------------------------------------------------------------------------------------
Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
3 consistent gets
0 physical reads
0 redo size
552 bytes sent via SQL*Net to client
552 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed
做一个简单的查询，可以看出有以下几点:
1全表扫描多了INMEMORY关健字;
2consistent gets急剧减少,少得有点可怜了，说明对内存访问极少;
3,cost是27相比传统的查询，看看in-memory快了多少，使用NO_INMEMORY hint让optimizer不走in-memory,而走传统的buffer cache
SQL> select /*+ NO_INMEMORY */ count(object_id) from t1;
COUNT(OBJECT_ID)
----------------
90905
Execution Plan
----------------------------------------------------------
Plan hash value: 3724264953
---------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
---------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 5 | 416 (1)| 00:00:01 |
| 1 | SORT AGGREGATE | | 1 | 5 | | |
| 2 | TABLE ACCESS FULL| T1 | 90905 | 443K| 416 (1)| 00:00:01 |
---------------------------------------------------------------------------
Statistics
----------------------------------------------------------
33 recursive calls
0 db block gets
1557 consistent gets
0 physical reads
0 redo size
552 bytes sent via SQL*Net to client
552 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
3 sorts (memory)
0 sorts (disk)
1 rows processed
传统的成本是416,consistent gets是1557,相比in-memory，访问数据大量增加.这是由于是传统是存储的是行式，而in-memory存放的是列式.速度快了非常多.有人可能会问，这里选择是单列，如果是count(*)了，性能还有这么明显吗?好，我们来看例子
SQL> select /*+ NO_INMEMORY */ count(*) from c##frank.t1;
COUNT(*)
----------
90905
Execution Plan
----------------------------------------------------------
Plan hash value: 3724264953
-------------------------------------------------------------------
| Id | Operation | Name | Rows | Cost (%CPU)| Time |
-------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 426 (1)| 00:00:01 |
| 1 | SORT AGGREGATE | | 1 | | |
| 2 | TABLE ACCESS FULL| T1 | 90905 | 426 (1)| 00:00:01 |
-------------------------------------------------------------------
Statistics
----------------------------------------------------------
5 recursive calls
0 db block gets
1536 consistent gets
1527 physical reads
0 redo size
544 bytes sent via SQL*Net to client
552 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed
在实验中还产生了物理读.再来看看in-memory的count(*)
SQL> select /*+ INMEMORY */ count(*) from c##frank.t1;
COUNT(*)
----------
90905
Execution Plan
----------------------------------------------------------
Plan hash value: 3724264953
----------------------------------------------------------------------------
| Id | Operation | Name | Rows | Cost (%CPU)| Time |
----------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 27 (0)| 00:00:01 |
| 1 | SORT AGGREGATE | | 1 | | |
| 2 | TABLE ACCESS INMEMORY FULL| T1 | 90905 | 27 (0)| 00:00:01 |
----------------------------------------------------------------------------
Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
3 consistent gets
0 physical reads
0 redo size
544 bytes sent via SQL*Net to client
552 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed
即使count(*),in-memory也和原来一样，快上很多倍.估计看到这样一对比很多人都已经坐不住了.
比如手动把表放到in-memory area中
SQL> create table t1(a int,b int);
Table created.
SQL> alter table t1 inmemory;
Table altered.
取消表在inmemory中存放
SQL> alter table t1 no inmemory;
Table altered.
还可以在栏位级定义哪一栏放到 in-memory area中
SQL> alter table t1 inmemory;
Table altered.
SQL> alter table t1 inmemory (a) no inmemory (b);
Table altered.
先定义整个表存放在in memory中,才能再定义某个栏位存放在in memory中.