oracle- insert性能优化

 看见朋友导入数据，花了很长时间都没完成！其实有很多快速的方法，整理下！

向表中插入数据有很多办法,但是方法不同，性能差别很大.


----1.原始语句
drop table t1 purge;

create table t1
(
sid number,
sname varchar2(20)
) tablespace test;



create or replace procedure proc01
(
sname varchar2
)
as
begin
    for i in 1..10000000
    loop
        execute immediate
        'insert into t1(sid,sname) values('||i||','''||sname||''')';
    commit;
    end loop;
end;
/



alter system flush shared_pool;

set timing on;

exec proc01('ocpyangtest');

已用时间:  02: 02: 54.12





----2.绑定变量



drop table t1 purge;

create table t1
(
sid number,
sname varchar2(20)
) tablespace test;


create or replace procedure proc02
(
sname varchar2
)
as
begin
    for i in 1..10000000
    loop
        execute immediate
    'insert into t1(sid,sname) values(:no'||','''||sname||''')' using i;
    commit;
    end loop;
end;
/



alter system flush shared_pool;

set timing on;

exec proc02('ocpyangtest');

已用时间:  00: 22: 59.79


select count(*) from t1;


----3.静态语句



drop table t1 purge;

create table t1
(
sid number,
sname varchar2(20)
) tablespace test;




create or replace procedure proc03
as
begin
    for i in 1..10000000
    loop

    insert into t1 values(i,'ocpyangtest');
    commit;
    end loop;
end;
/

alter system flush shared_pool;

set timing on;

exec proc03;

已用时间:  00: 20: 42.42

select count(*) from t1;






----4.批量提交


drop table t1 purge;

create table t1
(
sid number,
sname varchar2(20)
) tablespace test;




create or replace procedure proc04
as
begin
    for i in 1..10000000
    loop
    insert into t1 values(i,'ocpyangtest');
    end loop;
    commit;
end;
/


alter system flush shared_pool;

set timing on;

exec proc04;

已用时间:  00: 11: 48.42



----5.集合


drop table t2 purge;

create table t2
(
sid number,
sname varchar2(20)
) tablespace test;




alter system flush shared_pool;

select count(*) from t1;


set timing on;



insert into t2 select sid,sname from t1;


已用时间:  00: 01: 02.18

commit;


select count(*) from t2;



----6. 集合+append


drop table t2 purge;

create table t2
(
sid number,
sname varchar2(20)
) tablespace test;




alter system flush shared_pool;


select count(*) from t1;



set timing on;



insert /*  + append */ into t2 select sid,sname from t1;

已用时间:  00: 00: 36.94


commit;


select count(*) from t2;






----7. 集合+append+nologging


drop table t2 purge;

create table t2
(
sid number,
sname varchar2(20)
) nologging tablespace test;




alter system flush shared_pool;


select count(*) from t1;






set timing on;


insert /*  + append */ into t2 select sid ,sname from t1;

已用时间:  00: 00: 35.07

commit;


select count(*) from t2;






----8.数据加载1


drop table t2 purge;



alter system flush shared_pool;

select count(*) from t1;



set timing on;



create table t2
as
select sid,sname from t1;

已用时间:  00: 00: 25.91



select count(*) from t2;






----9.数据加载2


drop table t2 purge;




alter system flush shared_pool;

select count(*) from t1;



set timing on;



create table t2 nologging
as
select sid,sname from t1;

已用时间:  00: 00: 04.89



select count(*) from t2;



----10.数据加载+并行


drop table t2 purge;




alter system flush shared_pool;

select count(*) from t1;



set timing on;



create table t2 nologging parallel 24   --根据自己服务器情况
as
select sid,sname from t1;

已用时间:  00: 00: 02.89


select count(*) from t2;

 

转载于:https://blog.51cto.com/ocpyang/1194990