Dataguard broker setup and Standby Redo Logs
You wanna create an Oracle Dataguard configuration working in Maximum availability. This configuration uses of course standby redo log files. These files must be available on the primary and standby side (because of switchover : primary database becomes standby database). In your setup process you decided to create the standy redo log files before duplicating the primary database. After the creation of the Dataguard configuration, the standby database seems to miss the standby redo log files.
RFS[3]: No standby redo logfiles created
RFS[3]: Archived Log: '/u00/app/oracle/admin/DG/arch/DG_1_86_599565590.arc'
Below the description of the problem and the solution.
First of all you create the standby redo log files on the primary database :
alter database add standby logfile group 4
('/u00/oradata/DG/sredog4m1DG.dbf',
'/u01/oradata/DG/sredog4m2DG.dbf') SIZE 6M reuse;
alter database add standby logfile group 5
('/u00/oradata/DG/sredog5m1DG.dbf',
'/u01/oradata/DG/sredog5m2DG.dbf') SIZE 6M reuse;
alter database add standby logfile group 6
('/u00/oradata/DG/sredog6m1DG.dbf',
'/u01/oradata/DG/sredog6m2DG.dbf') SIZE 6M reuse;
alter database add standby logfile group 7
('/u00/oradata/DG/sredog7m1DG.dbf',
'/u01/oradata/DG/sredog7m2DG.dbf') SIZE 6M reuse;
Backup your primary database for standby with RMAN :
RMAN> connect target /
RMAN> backup incremental level 0 database;
RMAN> backup current controlfile for standby;
RMAN> backup archivelog all delete input;
SQL> startup nomount;
ORACLE instance started.
From the primary server, start RMAN to duplicate for standby :
RMAN> connect target /
connect connected to target database: DG (DBID=1512896598)
RMAN> connect auxiliary sys/manager@DG_SITE2
connected to auxiliary database: DG (not mounted)
RMAN> duplicate target database for standby dorecover nofilenamecheck;
After the duplicate verify the status on the standby database :
SQL> select database_role from v$database;
DATABASE_ROLE
----------------
PHYSICAL STANDBY
The standby redo log files have been correctly duplicated, but the files are NOT created on the filesystem.
SQL> select group# from v$standby_log;
GROUP#
----------
4
5
6
7
connect sys/manager
create configuration 'DG' as
primary database is 'DG_SITE1'
connect identifier is 'DG_SITE1.trivadistraining.com';
add database 'DG_SITE2' as
connect identifier is 'DG_SITE2.trivadistraining.com'
maintained as physical;
edit database 'DG_SITE1' set property StandbyArchiveLocation='/u00/app/oracle/admin/DG/arch';
edit database 'DG_SITE1' set property StandbyFileManagement='AUTO';
edit database 'DG_SITE1' set property LogXptMode='ARCH';
edit database 'DG_SITE2' set property StandbyArchiveLocation='/u00/app/oracle/admin/DG/arch';
edit database 'DG_SITE2' set property StandbyFileManagement='AUTO';
edit database 'DG_SITE2' set property LogXptMode='ARCH';
enable configuration;
show configuration;
On the standby side , following occured (extract of the alert.log file of the standby database) :
Completed: ALTER DATABASE RECOVER MANAGED STANDBY DATABASE THROUGH ALL SWITCHOVER DISCONNECT NODELAY
Sun Sep 23 14:52:54 2007
Redo Shipping Client Connected as PUBLIC
-- Connected User is Valid
RFS[3]: Assigned to RFS process 7535
RFS[3]: Identified database type as 'physical standby'
RFS[3]: No standby redo logfiles created
RFS[3]: Archived Log: '/u00/app/oracle/admin/DG/arch/DG_1_85_599565590.arc'
Sun Sep 23 14:52:55 2007
Media Recovery Log /u00/app/oracle/admin/DG/arch/DG_1_85_599565590.arc
Media Recovery Waiting for thread 1 sequence 86
Sun Sep 23 14:54:39 2007
RFS[3]: No standby redo logfiles created
RFS[3]: Archived Log: '/u00/app/oracle/admin/DG/arch/DG_1_86_599565590.arc'
RFS[3]: No standby redo logfiles created
RFS[3]: Archived Log: '/u00/app/oracle/admin/DG/arch/DG_1_87_599565590.arc'
Sun Sep 23 14:54:40 2007
Media Recovery Log /u00/app/oracle/admin/DG/arch/DG_1_86_599565590.arc
Media Recovery Log /u00/app/oracle/admin/DG/arch/DG_1_87_599565590.arc
After having enabled the Dataguard configuration, the standby redo log disappeared :
SQL> select group# from v$standby_log;
no rows selected
The standby database complains in the alert.log file :
RFS[3]: No standby redo logfiles created
RFS[3]: Archived Log: '/u00/app/oracle/admin/DG/arch/DG_1_86_599565590.arc'
On the primary database (in dgmgrl) :
DGMGRL> edit database 'DG_SITE2' set state = 'LOG-APPLY-OFF';
Succeeded.
SQL> @create_standy_log.sql
Database altered.
Database altered.
Database altered.
Database altered.
DGMGRL> edit database 'DG_SITE2' set state = 'ONLINE';
Succeeded.
SQL> alter system archive log current;
System altered.
On the standby side , the database doesn't complain anymore and uses the standby redo log files :
RFS[5]: Successfully opened standby log 4: '/u00/oradata/DG/sredog4m1DG.dbf'
Sun Sep 23 14:57:28 2007
Media Recovery Log /u00/app/oracle/admin/DG/arch/DG_1_91_599565590.arc
Media Recovery Waiting for thread 1 sequence 92