关于物理DG的扩展表空间，首先要看备库端的参数standby_file_management,该参数可以设置为手动管理(MANUAL)和自动管理(AUTO).通常该参数配置为AUTO，在自动管理模式下，当表空间大小在主库端变更时(不论是alter tablespace add datafile还是alter database datafile *** resize)，备库能够自动侦测并调整大小，不需要手动干预。以下是物理dataguard扩展表空间的例子:
oracle 10.2.0.1，linux上的物理DG，首先standby处于日志接受模式：
SQL> select instance_name ,status from v$instance;
INSTANCE_NAME STATUS
-------------------------------- ------------------------
orcl MOUNTED
SQL>
在主库端查看归档日志序号：
SQL> select max(sequence#) from v$archived_log;
MAX(SEQUENCE#)
--------------
5593
SQL>
在备库端查看日志应用序号：
select sequence#,applied from v$archived_log;
5583 YES
5584 YES
SEQUENCE# APPLIE
---------- ------
5585 YES
5586 YES
5587 YES
5588 YES
5589 YES
5590 YES
5591 YES
5592 YES
5593 YES
2528 rows selected.
SQL>
日志应用正常。
在主库端查看表空间使用率如下：
SQL> select t.tablespace_name,t.total "总大小(M)",t.total-f.free "已使用(M)",f.free "剩余容量(M)",
(t.total-f.free)/t.total*100 "使用率%" from
(select tablespace_name,sum(bytes)/1024/1024 total
2 3 4 from dba_data_files group by tablespace_name) t,
5 (select tablespace_name,sum(bytes)/1024/1024 free
6 from dba_free_space group by tablespace_name) f
7 where t.tablespace_name=f.tablespace_name(+)
8 ;
TABLESPACE_NAME 总大尣?M) 已使産?M) 剩余容量(M) 使用玣?
-------------------- ------------ ------------ --------------- ----------
APPS_DATA 7524 794.6875 6729.3125 10.5620348
HALFENTERSEND 1024 59.1875 964.8125 5.7800293
UNDOTBS1 10240 78.5625 10161.4375 .767211914
SYSAUX 570 528.375 41.625 92.6973684
BSKILL 1024 183.0625 840.9375 17.8771973
HALFENTER 3118 2044.5625 1073.4375 65.5728833
USERS 872.5 64.5 808 7.39255014
DEVELOPE 11754 2056.5 9697.5 17.4961715
STATE_DATA 3072 2157.8125 914.1875 70.2412923
SYSTEM 960 582.9375 377.0625 60.7226563
SKILLS_DATA 13568 12982.4375 585.5625 95.6842386
TABLESPACE_NAME 总大尣?M) 已使産?M) 剩余容量(M) 使用玣?
-------------------- ------------ ------------ --------------- ----------
MAILCHECK_WORK 256 48.0625 207.9375 18.7744141
IMPORTUSER 1024 799.1875 224.8125 78.0456543
PUBCODE_DATA 512 253.875 258.125 49.5849609
EMAIL_WORK 11744 9579.0625 2164.9375 81.5655867
EMPLY 1024 863.1875 160.8125 84.2956543
EMPLYSEND 1024 11.1875 1012.8125 1.0925293
17 rows selected.
SQL>
这里SKILLS_DATA表空间13568M，使用率达到了95.6842386%。
在备库端查询:
SQL> alter database recover managed standby database cancel;
Database altered.
SQL> alter database open read only;
Database altered.
SQL>
SQL> select t.tablespace_name,t.total "总大小(M)",t.total-f.free "已使用(M)",f.free "剩余容量(M)",
2 (t.total-f.free)/t.total*100 "使用率%" from
3 (select tablespace_name,sum(bytes)/1024/1024 total
4 from dba_data_files group by tablespace_name) t,
5 (select tablespace_name,sum(bytes)/1024/1024 free
6 from dba_free_space group by tablespace_name) f
7 where t.tablespace_name=f.tablespace_name(+)
8 ;
TABLESPACE_NAME 总大尣?M) 已使産?M) 剩余容量(M) 使用玣?
-------------------- ------------ ------------ --------------- ----------
APPS_DATA 7524 794.6875 6729.3125 10.5620348
HALFENTERSEND 1024 59.1875 964.8125 5.7800293
UNDOTBS1 10240 78.5625 10161.4375 .767211914
SYSAUX 570 528.375 41.625 92.6973684
BSKILL 1024 183.0625 840.9375 17.8771973
HALFENTER 3118 2044.5625 1073.4375 65.5728833
USERS 872.5 64.5 808 7.39255014
DEVELOPE 11754 2056.5 9697.5 17.4961715
STATE_DATA 3072 2157.8125 914.1875 70.2412923
SYSTEM 960 582.9375 377.0625 60.7226563
SKILLS_DATA 13568 12982.4375 585.5625 95.6842386
TABLESPACE_NAME 总大尣?M) 已使産?M) 剩余容量(M) 使用玣?
-------------------- ------------ ------------ --------------- ----------
MAILCHECK_WORK 256 48.0625 207.9375 18.7744141
IMPORTUSER 1024 799.1875 224.8125 78.0456543
PUBCODE_DATA 512 253.875 258.125 49.5849609
EMAIL_WORK 11744 9579.0625 2164.9375 81.5655867
EMPLY 1024 863.1875 160.8125 84.2956543
EMPLYSEND 1024 11.1875 1012.8125 1.0925293
17 rows selected.
SQL>
也是完全一致的。
我的环境里，SKILLS_DATA表空间有两个数据文件，但是在主库端和备库端的目录不同:
primary:
SQL> select file_name from dba_data_files where tablespace_name='SKILLS_DATA';
FILE_NAME
--------------------------------------------------------------------------------
/u01/oracle/oradata/research/skills_data.dbf
/u01/oracle/oradata/research/skills_data01.dbf
standby:
SQL> select file_name from dba_data_files where tablespace_name='SKILLS_DATA';
FILE_NAME
--------------------------------------------------------------------------------
/opt/oracle/oradata/research/skills_data.dbf
/opt/oracle/oradata/research/skills_data01.dbf
将备库切换回日志接收模式：
SQL> alter database recover managed standby database disconnect from session;
Database altered.
SQL>
在primary:
SQL> alter tablespace SKILLS_DATA add datafile '/u01/oracle/oradata/research/skills_data02' size 8G;
primary:
SQL> select file_name from dba_data_files where tablespace_name='SKILLS_DATA';
FILE_NAME
--------------------------------------------------------------------------------
/u01/oracle/oradata/research/skills_data.dbf
/u01/oracle/oradata/research/skills_data01.dbf
/u01/oracle/oradata/research/skills_data02
SQL> select t.tablespace_name,t.total "总大小(M)",t.total-f.free "已使用(M)",f.free "剩余容量(M)",
2 (t.total-f.free)/t.total*100 "使用率%" from
3 (select tablespace_name,sum(bytes)/1024/1024 total
4 from dba_data_files group by tablespace_name) t,
5 (select tablespace_name,sum(bytes)/1024/1024 free
6 from dba_free_space group by tablespace_name) f
7 where t.tablespace_name=f.tablespace_name(+)
8 ;
TABLESPACE_NAME 总大尣?M) 已使産?M) 剩余容量(M) 使用玣?
-------------------- ------------ ------------ --------------- ----------
APPS_DATA 7524 794.6875 6729.3125 10.5620348
HALFENTERSEND 1024 59.1875 964.8125 5.7800293
UNDOTBS1 10240 79.5625 10160.4375 .776977539
SYSAUX 570 528.375 41.625 92.6973684
BSKILL 1024 183.0625 840.9375 17.8771973
HALFENTER 3118 2044.5625 1073.4375 65.5728833
USERS 872.5 64.5 808 7.39255014
DEVELOPE 11754 2056.5 9697.5 17.4961715
STATE_DATA 3072 2157.8125 914.1875 70.2412923
SYSTEM 960 582.9375 377.0625 60.7226563
SKILLS_DATA 21760 12982.5 8777.5 59.6622243
TABLESPACE_NAME 总大尣?M) 已使産?M) 剩余容量(M) 使用玣?
-------------------- ------------ ------------ --------------- ----------
MAILCHECK_WORK 256 48.0625 207.9375 18.7744141
IMPORTUSER 1024 799.1875 224.8125 78.0456543
PUBCODE_DATA 512 253.875 258.125 49.5849609
EMAIL_WORK 11744 9579.0625 2164.9375 81.5655867
EMPLY 1024 863.1875 160.8125 84.2956543
EMPLYSEND 1024 11.1875 1012.8125 1.0925293
17 rows selected.
SQL>
standby:
SQL> alter database recover managed standby database cancel;
Database altered.
SQL> alter datbase open read only;
alter datbase open read only
*
ERROR at line 1:
ORA-00940: invalid ALTER command
SQL> alter database open read only;
Database altered.
SQL>
standby:
SQL> select file_name from dba_data_files where tablespace_name='SKILLS_DATA';
FILE_NAME
--------------------------------------------------------------------------------
/opt/oracle/oradata/research/skills_data.dbf
/opt/oracle/oradata/research/skills_data01.dbf
SQL>
standby 切回日志接收模式：
primary:
SQL> alter system switch logfile;
System altered.
SQL> /
System altered.
SQL>
primary 查询：
SQL> select max(sequence#) from v$archived_log;
MAX(SEQUENCE#)
--------------
5598
SQL>
standby端切换到open read only状态，然后再次查询：
standby查询：
5588 YES
5589 YES
SEQUENCE# APPLIE
---------- ------
5590 YES
5591 YES
5592 YES
5593 YES
5594 YES
5595 YES
5596 YES
5597 YES
5598 YES
2528 rows selected.
SQL>
然后在standby端查询:
SQL> select file_name from dba_data_files where tablespace_name='SKILLS_DATA';
FILE_NAME
--------------------------------------------------------------------------------
/opt/oracle/oradata/research/skills_data.dbf
/opt/oracle/oradata/research/skills_data01.dbf
/u01/oracle/oradata/research/skills_data02
SQL>
我们可以看到，standby已经同步看到了新加的数据文件。
以上是standby_file_management为AUTO的方式，另外，如果数据文件在裸设备下面，一定要先在主库端和备库端都建立好裸设备，尤其是备库端，不然备库端无法找到相应的裸设备就会出问题。如果不是AUTO而是MANUAL，则需要手动的在主库和备库端同时修改。