Oracle数据泵导入的时候创建索引是否会使用并行？

时间：2021-09-18 04:25:16

一、疑问，Oracle数据泵导入的时候创建索引是否会使用并行？

某客户需要使用数据泵进行迁移，客户咨询导入的时间能不能加快一点。

那么如何加快导入的速度呢？ 多加一些并行，那么创建索引内部的索引并行度是否会调整呢？

带着这些疑问看看Oracle数据泵并行参数与导入后创建索引的并行度是否有关系！

二、实验测试

2.1测试数据准备

Oracel11.2.0.4

--分区表创建

create user yz identified by yz;

grant dba to yz;

conn yz/yz

create table a1(id number,

    deal_date date, area_code number, contents varchar2(4000))

 partition by range(deal_date)

 (

 partition p1 values less than(to_date('2019-02-01','yyyy-mm-dd')),

 partition p2 values less than(to_date('2019-03-01','yyyy-mm-dd')),

 partition p3 values less than(to_date('2019-04-01','yyyy-mm-dd')),

 partition p4 values less than(to_date('2019-05-01','yyyy-mm-dd')),

 partition p5 values less than(to_date('2019-06-01','yyyy-mm-dd')),

 partition p6 values less than(to_date('2019-07-01','yyyy-mm-dd')),

 partition p7 values less than(to_date('2019-08-01','yyyy-mm-dd')),

 partition p8 values less than(to_date('2019-09-01','yyyy-mm-dd')),

 partition p9 values less than(to_date('2019-10-01','yyyy-mm-dd')),

 partition p10 values less than(to_date('2019-11-01','yyyy-mm-dd')),

 partition p11 values less than(to_date('2019-12-01','yyyy-mm-dd')),

 partition p12 values less than(to_date('2020-01-01','yyyy-mm-dd')),

 partition p13 values less than(to_date('2020-02-01','yyyy-mm-dd')),

 partition p14 values less than(to_date('2020-03-01','yyyy-mm-dd')),

 partition p15 values less than(to_date('2020-04-01','yyyy-mm-dd')),

 partition p16 values less than(to_date('2020-05-01','yyyy-mm-dd')),

 partition p17 values less than(to_date('2020-06-01','yyyy-mm-dd')),

 partition p18 values less than(to_date('2020-07-01','yyyy-mm-dd')),

 partition p19 values less than(to_date('2020-08-01','yyyy-mm-dd')),

 partition p20 values less than(to_date('2020-09-01','yyyy-mm-dd')),

 partition p31 values less than(to_date('2020-10-01','yyyy-mm-dd')),

 partition p32 values less than(to_date('2020-11-01','yyyy-mm-dd')),

 partition p33 values less than(to_date('2020-12-01','yyyy-mm-dd')),

 partition p34 values less than(to_date('2021-01-01','yyyy-mm-dd')),

 partition p35 values less than(to_date('2021-02-01','yyyy-mm-dd')),

 partition p36 values less than(to_date('2021-03-01','yyyy-mm-dd')),

 partition p37 values less than(to_date('2021-04-01','yyyy-mm-dd')),

 partition p38 values less than(to_date('2021-05-01','yyyy-mm-dd')),

 partition p39 values less than(to_date('2021-06-01','yyyy-mm-dd')),

 partition p40 values less than(to_date('2021-07-01','yyyy-mm-dd'))

 );

insert into a1 (id,deal_date,area_code,contents)

 select rownum,

          to_date(to_char(sysdate-900,'J')+ trunc(dbms_random.value(0,200)),'J'),

ceil(dbms_random.value(590,599)),

 rpad('*',400,'*')

 from dual

 connect by rownum <= 100000;

commit;

create table a2(id number,

    deal_date date, area_code number, contents varchar2(4000))

 partition by range(deal_date)

 (

 partition p1 values less than(to_date('2019-02-01','yyyy-mm-dd')),

 partition p2 values less than(to_date('2019-03-01','yyyy-mm-dd')),

 partition p3 values less than(to_date('2019-04-01','yyyy-mm-dd')),

 partition p4 values less than(to_date('2019-05-01','yyyy-mm-dd')),

 partition p5 values less than(to_date('2019-06-01','yyyy-mm-dd')),

 partition p6 values less than(to_date('2019-07-01','yyyy-mm-dd')),

 partition p7 values less than(to_date('2019-08-01','yyyy-mm-dd')),

 partition p8 values less than(to_date('2019-09-01','yyyy-mm-dd')),

 partition p9 values less than(to_date('2019-10-01','yyyy-mm-dd')),

 partition p10 values less than(to_date('2019-11-01','yyyy-mm-dd')),

 partition p11 values less than(to_date('2019-12-01','yyyy-mm-dd')),

 partition p12 values less than(to_date('2020-01-01','yyyy-mm-dd')),

 partition p13 values less than(to_date('2020-02-01','yyyy-mm-dd')),

 partition p14 values less than(to_date('2020-03-01','yyyy-mm-dd')),

 partition p15 values less than(to_date('2020-04-01','yyyy-mm-dd')),

 partition p16 values less than(to_date('2020-05-01','yyyy-mm-dd')),

 partition p17 values less than(to_date('2020-06-01','yyyy-mm-dd')),

 partition p18 values less than(to_date('2020-07-01','yyyy-mm-dd')),

 partition p19 values less than(to_date('2020-08-01','yyyy-mm-dd')),

 partition p20 values less than(to_date('2020-09-01','yyyy-mm-dd')),

 partition p31 values less than(to_date('2020-10-01','yyyy-mm-dd')),

 partition p32 values less than(to_date('2020-11-01','yyyy-mm-dd')),

 partition p33 values less than(to_date('2020-12-01','yyyy-mm-dd')),

 partition p34 values less than(to_date('2021-01-01','yyyy-mm-dd')),

 partition p35 values less than(to_date('2021-02-01','yyyy-mm-dd')),

 partition p36 values less than(to_date('2021-03-01','yyyy-mm-dd')),

 partition p37 values less than(to_date('2021-04-01','yyyy-mm-dd')),

 partition p38 values less than(to_date('2021-05-01','yyyy-mm-dd')),

 partition p39 values less than(to_date('2021-06-01','yyyy-mm-dd')),

 partition p40 values less than(to_date('2021-07-01','yyyy-mm-dd'))

 );

insert into a2 (id,deal_date,area_code,contents)

 select rownum,

          to_date(to_char(sysdate-900,'J')+ trunc(dbms_random.value(0,200)),'J'),

ceil(dbms_random.value(590,599)),

 rpad('*',400,'*')

 from dual

 connect by rownum <= 200000;

commit;

alter table a1 add constraint pk_id primary key (id);

alter table a2 add constraint pk_id_time primary key(id,deal_date);

SQL> create index cc_id on a1(id);

create index cc_id on a1(id)

                         *

ERROR at line 1:

ORA-01408: such column list already indexed

SQL> select index_name,status from user_indexes where table_name in('A1','A2');

INDEX_NAME                     STATUS

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

PK_ID_TIME                     VALID

PK_ID                          VALID    

Alter table a1 drop constraint pk_id;

Alter table a2 drop constraint pk_id_time;

create index cc_id on a1(id) LOCAL;

alter table a1 add constraint pk_id primary key (id) USING INDEX cc_id ;

ORA-14196: Specified index cannot be used to enforce the constraint.

DROP INDEX CC_ID;

create index cc_id on a1(id) ;

alter table a1 add constraint pk_id primary key (id) USING INDEX cc_id ;

create index cc_id_DATE on a2(id,DEAL_DATE) LOCAL;

alter table a2 add constraint pk_id_DATE primary key (id,DEAL_DATE) USING INDEX cc_id_DATE ;

https://www.cnblogs.com/lvcha001/p/10218318.html

索引可以认为分3种，非分区索引，全局XX索引，可以是全局范围分区、全局哈希分区，这种情况会根据规则将数据打散，而不是根据实际表的数据进行打散！
  本地索引，完全根据分区表的子分区，一个子分区一个索引！但是我们观察重建索引其实是一样的操作，本次测试为了后续重建索引，创建不同的分区类型，非分区索引使用主键！

create index ind_hash on a1(id,0) global partition by hash (id) partitions 8 online;

SQL>  select index_name,status from user_indexes where table_name in('A1','A2');

INDEX_NAME                     STATUS

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

CC_ID                          VALID

CC_ID_DATE                     N/A

IND_HASH                       N/A

select index_name,PARTITION_NAME,HIGH_VALUE,STATUS,TABLESPACE_NAME from dba_ind_partitions
 where index_owner='YZ' and index_name IN('CC_ID_DATE','IND_HASH');

2.2 导入SQL文件测试

nohup time expdp \'/ as sysdba\' directory=dump dumpfile=yang%u.dmp logfile=yang.log tables=yz.a1,yz.a2 FLASHBACK_SCN=1017463 parallel=2 &

情况一、导出并行度2，导入并行度2，观察SQL脚本

nohup time impdp \'/ as sysdba\' directory=dump dumpfile=yang%u.dmp logfile=yang.log tables=yz.a1,yz.a2 parallel=2 sqlfile=table01.sql &
Processing object type TABLE_EXPORT/TABLE/PROCACT_INSTANCE

Processing object type TABLE_EXPORT/TABLE/TABLE

Processing object type TABLE_EXPORT/TABLE/INDEX/INDEX

Processing object type TABLE_EXPORT/TABLE/INDEX/FUNCTIONAL_INDEX/INDEX

Processing object type TABLE_EXPORT/TABLE/CONSTRAINT/CONSTRAINT

Processing object type TABLE_EXPORT/TABLE/INDEX/STATISTICS/INDEX_STATISTICS

Processing object type TABLE_EXPORT/TABLE/INDEX/STATISTICS/FUNCTIONAL_INDEX/INDEX_STATISTICS

Job "SYS"."SYS_SQL_FILE_TABLE_01" successfully completed at Wed Aug 11 07:00:04 2021 elapsed 0 00:00:03

-- new object type path: TABLE_EXPORT/TABLE/TABLE

-- CONNECT SYS

CREATE TABLE "YZ"."A2"

   (    "ID" NUMBER,

······

-- new object type path: TABLE_EXPORT/TABLE/INDEX/INDEX

-- CONNECT YZ

CREATE INDEX "YZ"."CC_ID_DATE" ON "YZ"."A2" ("ID", "DEAL_DATE")

  PCTFREE 10 INITRANS 2 MAXTRANS 255

  STORAGE(

  BUFFER_POOL DEFAULT FLASH_CACHE DEFAULT CELL_FLASH_CACHE DEFAULT) LOCAL

 (PARTITION "P1"

  PCTFREE 10 INITRANS 2 MAXTRANS 255 LOGGING

  STORAGE(

  BUFFER_POOL DEFAULT FLASH_CACHE DEFAULT CELL_FLASH_CACHE DEFAULT)

  TABLESPACE "USERS" ,

 PARTITION "P2"

······

 PARTITION "P40"

  PCTFREE 10 INITRANS 2 MAXTRANS 255 LOGGING

  STORAGE(

  BUFFER_POOL DEFAULT FLASH_CACHE DEFAULT CELL_FLASH_CACHE DEFAULT)

  TABLESPACE "USERS" ) PARALLEL 1 ;

  ALTER INDEX "YZ"."CC_ID_DATE" NOPARALLEL;

CREATE INDEX "YZ"."CC_ID" ON "YZ"."A1" ("ID")

  PCTFREE 10 INITRANS 2 MAXTRANS 255

  STORAGE(INITIAL 65536 NEXT 1048576 MINEXTENTS 1 MAXEXTENTS 2147483645

  PCTINCREASE 0 FREELISTS 1 FREELIST GROUPS 1

  BUFFER_POOL DEFAULT FLASH_CACHE DEFAULT CELL_FLASH_CACHE DEFAULT)

  TABLESPACE "USERS" PARALLEL 1 ;

  ALTER INDEX "YZ"."CC_ID" NOPARALLEL;

-- new object type path: TABLE_EXPORT/TABLE/INDEX/FUNCTIONAL_INDEX/INDEX

CREATE INDEX "YZ"."IND_HASH" ON "YZ"."A1" ("ID", 0)

  PCTFREE 10 INITRANS 2 MAXTRANS 255

  STORAGE(

  BUFFER_POOL DEFAULT FLASH_CACHE DEFAULT CELL_FLASH_CACHE DEFAULT)

  TABLESPACE "USERS"  GLOBAL PARTITION BY HASH ("ID")

 (PARTITION "SYS_P41"

  TABLESPACE "USERS" ,

 PARTITION "SYS_P42"

  TABLESPACE "USERS" ,

 PARTITION "SYS_P43"

  TABLESPACE "USERS" ,

 PARTITION "SYS_P44"

  TABLESPACE "USERS" ,

 PARTITION "SYS_P45"

  TABLESPACE "USERS" ,

 PARTITION "SYS_P46"

  TABLESPACE "USERS" ,

 PARTITION "SYS_P47"

  TABLESPACE "USERS" ,

 PARTITION "SYS_P48"

  TABLESPACE "USERS" ) PARALLEL 1 ;

  ALTER INDEX "YZ"."IND_HASH" NOPARALLEL;

-- new object type path: TABLE_EXPORT/TABLE/CONSTRAINT/CONSTRAINT

-- CONNECT SYS

ALTER TABLE "YZ"."A1" ADD CONSTRAINT "PK_ID" PRIMARY KEY ("ID")

  USING INDEX "YZ"."CC_ID"  ENABLE;

ALTER TABLE "YZ"."A1" ADD SUPPLEMENTAL LOG GROUP "GGS_87350" ("ID") ALWAYS;

ALTER TABLE "YZ"."A1" ADD SUPPLEMENTAL LOG DATA (PRIMARY KEY) COLUMNS;

ALTER TABLE "YZ"."A1" ADD SUPPLEMENTAL LOG DATA (UNIQUE INDEX) COLUMNS;

ALTER TABLE "YZ"."A1" ADD SUPPLEMENTAL LOG DATA (FOREIGN KEY) COLUMNS;

ALTER TABLE "YZ"."A2" ADD CONSTRAINT "PK_ID_DATE" PRIMARY KEY ("ID", "DEAL_DATE")

  USING INDEX "YZ"."CC_ID_DATE"  ENABLE;

ALTER TABLE "YZ"."A2" ADD SUPPLEMENTAL LOG GROUP "GGS_87381" ("ID", "DEAL_DATE") ALWAYS;

ALTER TABLE "YZ"."A2" ADD SUPPLEMENTAL LOG DATA (PRIMARY KEY) COLUMNS;

ALTER TABLE "YZ"."A2" ADD SUPPLEMENTAL LOG DATA (UNIQUE INDEX) COLUMNS;

ALTER TABLE "YZ"."A2" ADD SUPPLEMENTAL LOG DATA (FOREIGN KEY) COLUMNS;

-- new object type path: TABLE_EXPORT/TABLE/INDEX/STATISTICS/INDEX_STATISTICS

······

并行度1！
情况二、导出并行度2，导入并行度4，观察SQL脚本

$ nohup time impdp \'/ as sysdba\' directory=dump dumpfile=yang%u.dmp logfile=yang.log tables=yz.a1,yz.a2 parallel=4 sqlfile=table02.sql &

$ cat dump/table02.sql |grep PARALLEL

TABLESPACE "USERS" ) PARALLEL 1 ;

ALTER INDEX "YZ"."CC_ID_DATE" NOPARALLEL;

TABLESPACE "USERS" PARALLEL 1 ;

ALTER INDEX "YZ"."CC_ID" NOPARALLEL;

TABLESPACE "USERS" ) PARALLEL 1 ;

ALTER INDEX "YZ"."IND_HASH" NOPARALLEL;

通过测试，我们可以得知数据泵导入创建索引并行度就是1！！！  这种情况除非数据库参数设置对象是AUTO选择并行，如果是Manual的情况则无法使用并行加快速度！

2.3 如何对数据泵导入使用并行创建索引加快速度！

参考

https://blog.51cto.com/wyzwl/2333565

？为什么脚本要排除约束？    感兴趣的小伙伴可以测试一下。
目标端创建用户、授权后，导入表数据！

*****************************************************数据导入**************************************

cat >imp_data.par <<EOF
userid='/ as sysdba'
directory=dump
dumpfile=yang%u.dmp
logfile=imp_data.log
cluster=no
parallel=2
exclude= index,constraint
EOF
--排除索引和约束，执行导入
nohup impdp parfile=imp_data.par > imp_data.out &

*****************************************************索引及约束导入**************************************

--通过sqlfile参数生成创建索引语句
cat >imp_ind_con.par <<EOF
userid='/ as sysdba'
directory=dump
dumpfile=yang%u.dmp
sqlfile=imp_ind_con.sql
logfile=imp_ind_con.log
cluster=no
parallel=2
tables=yz.a1,yz.a2
include=index,constraint
EOF

--执行生成创建索引语句（实际并不会导入）
nohup impdp parfile= imp_ind_con.par > imp_ind_con.out &
--修改创建索引的并行度，并行度建议不超过CPU核数的1.5倍
--LINUX环境使用
sed -i 's/PARALLEL 1/PARALLEL 16/g' imp_ind_con.sql
--因AIX环境sed没有-i参数，可以使用如下两种方法：
perl -pi -e 's/ PARALLEL 1/PARALLEL 16/g' imp_ind_con.sql
或者
vi imp_ind_con.sql << EOF
:%s/ PARALLEL 1/PARALLEL 16/g
:wq
EOF

*****************************************************替换效果***************************************

[oracle@t2 dump]$ cat imp_ind_con.sql|grep PARALLEL
TABLESPACE "USERS" ) PARALLEL 1 ;
ALTER INDEX "YZ"."CC_ID_DATE" NOPARALLEL;
TABLESPACE "USERS" PARALLEL 1 ;
ALTER INDEX "YZ"."CC_ID" NOPARALLEL;
TABLESPACE "USERS" ) PARALLEL 1 ;
ALTER INDEX "YZ"."IND_HASH" NOPARALLEL;
[oracle@t2 dump]$ sed -i 's/PARALLEL 1/PARALLEL 16/g' imp_ind_con.sql
[oracle@t2 dump]$ cat imp_ind_con.sql|grep PARALLEL
TABLESPACE "USERS" ) PARALLEL 16 ;
ALTER INDEX "YZ"."CC_ID_DATE" NOPARALLEL;
TABLESPACE "USERS" PARALLEL 16 ;
ALTER INDEX "YZ"."CC_ID" NOPARALLEL;
TABLESPACE "USERS" ) PARALLEL 16 ;
ALTER INDEX "YZ"."IND_HASH" NOPARALLEL;

***************************************************************************************************

$more 观察SQL脚本

-- new object type path: TABLE_EXPORT/TABLE/INDEX/INDEX
-- CONNECT YZ
CREATE INDEX "YZ"."CC_ID_DATE" ON "YZ"."A2" ("ID", "DEAL_DATE")
PCTFREE 10 INITRANS 2 MAXTRANS 255
······
-- new object type path: TABLE_EXPORT/TABLE/CONSTRAINT/CONSTRAINT
-- CONNECT SYS
ALTER TABLE "YZ"."A1" ADD CONSTRAINT "PK_ID" PRIMARY KEY ("ID")
USING INDEX "YZ"."CC_ID" ENABLE;

先创建索引在整约束！

--等导入完数据之后，执行建索引的SQL：
$vi imp_ind_con.sh
sqlplus / as sysdba <<EOF
set timing on
set echo on
set verify on
spool imp_ind_con.log
@imp_ind_con.sql
spool off
exit
EOF
--执行建索引的SQL
nohup sh imp_ind_con.sh> imp_ind_con.out &

疑问一、导出表的dump有创建用户的语句吗？ 如何导出创建用户的SQL语句

--从只导出表的dump，导入create user 提示Not found user

$ nohup time impdp \'/ as sysdba\' directory=dump dumpfile=yang%u.dmp logfile=yang.log include=user parallel=1 sqlfile=user01.sql &

ORA-39168: Object path USER was not found.

--从导出整个schema dump再次测试！

$nohup time expdp \'/ as sysdba\' directory=dump dumpfile=yanga%u.dmp logfile=yang.log SCHEMAS=yz parallel=2

$ scp /home/oracle/script/dump/yanga*.dmp t2:/home/oracle/script/dump/.

$ nohup time impdp \'/ as sysdba\' directory=dump dumpfile=yanga%u.dmp logfile=yang.log include=user parallel=1 sqlfile=user02.sql &

$ cat user02.sql

-- CONNECT SYSTEM

 CREATE USER "YZ" IDENTIFIED BY VALUES 'S:C9A5297B9802EBB85A3BE800929ECE1BFCCB00146E58E0FBB055A937869F;86EF13A1088170F5'

      DEFAULT TABLESPACE "USERS"

      TEMPORARY TABLESPACE "TEMP";

疑问二、数据泵导出用户后，导入这个用户，这个用户之前在DB不存在，Oracle会自动创建这个用户吗？

 Starting "SYS"."SYS_EXPORT_SCHEMA_01":  "/******** AS SYSDBA" directory=dump dumpfile=yanga%u.dmp logfile=yang.log SCHEMAS=yz parallel=2 
nohup time impdp \'/ as sysdba\' directory=dump dumpfile=yanga%u.dmp logfile=yanga.log parallel=4 &

SQL> select username,account_status from dba_users where username='YZ';

no rows selected

SQL>
SQL> r
1* select username,account_status from dba_users where username='YZ'

USERNAME ACCOUNT_STATUS
------------------------------ --------------------------------
YZ OPEN

是可以自动创建用户的！

Processing object type SCHEMA_EXPORT/USER
Processing object type SCHEMA_EXPORT/SYSTEM_GRANT
Processing object type SCHEMA_EXPORT/ROLE_GRANT
Processing object type SCHEMA_EXPORT/DEFAULT_ROLE
Processing object type SCHEMA_EXPORT/PRE_SCHEMA/PROCACT_SCHEMA
Processing object type SCHEMA_EXPORT/TABLE/PROCACT_INSTANCE
Processing object type SCHEMA_EXPORT/TABLE/TABLE
Processing object type SCHEMA_EXPORT/TABLE/TABLE_DATA
. . imported "YZ"."A2":"P3" 12.53 MB 30997 rows

. . imported "YZ"."A2":"P40" 0 KB 0 rows
Processing object type SCHEMA_EXPORT/TABLE/INDEX/INDEX
Processing object type SCHEMA_EXPORT/TABLE/INDEX/FUNCTIONAL_INDEX/INDEX
Processing object type SCHEMA_EXPORT/TABLE/CONSTRAINT/CONSTRAINT
Processing object type SCHEMA_EXPORT/TABLE/INDEX/STATISTICS/INDEX_STATISTICS
Processing object type SCHEMA_EXPORT/TABLE/INDEX/STATISTICS/FUNCTIONAL_INDEX/INDEX_STATISTICS
Job "SYS"."SYS_IMPORT_FULL_01" successfully completed at Wed Aug 11 12:47:16 2021 elapsed 0 00:00:17

相关文章