Using Guaranteed Restore Points for Primary and FailoverSteps to use Guaranteed Restore Points for backouts of database changes in a dataguard broker environmentFailoverCreate the restore point on the failover database first, if there is addtional standbys do the same on each standbySQLALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;SQLcreate restore point AT_test GUARANTEE FLASHBACK DATABASE;SQLALTER DATABASE RECOVER MANAGED STANDBY DATABASE THROUGH ALL SWITCHOVER DISCONNECT USING CURRENT LOGFILE;PrimaryCreate the restore point on the primary databaseSQLalter system switch logfile;SQLalter system switch logfile;SQLalter system switch logfile;SQLcreate restore point AT_test GUARANTEE FLASHBACK DATABASE;Back out steps FlashbackIn the event of having to backout us the following steps to flashback to the restore pointsDisabling the configuration in dataguard broker Disable the configuration in dataguard brokerDGMGRL show configuration check no errors before starting thisDGMGRL disable FAST_START FAILOVER; Only needed if fast start failover is enabledDGMGRL DISABLE CONFIGURATION;PrimaryFlashback the primary database to the restore pointSQLshu immediateSQLstartup mountSQLFLASHBACK DATABASE TO RESTORE POINT AT_test;SQLalter database open resetlogs;StandbyFlashback the primary database to the restore pointSQLshu immediateSQLstartup mountSQLFLASHBACK DATABASE TO RESTORE POINT AT_test;Enabling the configuration in dataguard broker Enable the configuration in dataguard brokerDGMGRL show configuration check no errors before starting thisDGMGRL enable CONFIGURATION;DGMGRL enable FAST_START FAILOVER; Only needed if fast start failover is enabledPost Flashback ChecksCheck the standby alert log to make sure the logs are being applied, you should see output like belowMedia Recovery Waiting for thread 1 sequence 8 (in transit)Recovery of Online Redo Log: Thread 1 Group 13 Seq 8 Reading mem 0Mem# 0: /wload/test2/app/oracle/u02/oradata/testdb1/srl13a.rdoMem# 1: /wload/test2/app/oracle/u03/oradata/testdb1/srl13b.rdoRFS3: Selected log 14 for thread 1 sequence 9 dbid 528024012 branch 780852273Tue Apr 17 15:27:39 2012Archived Log entry 997 added for thread 1 sequence 8 ID 0x1fc46837 dest 1:Media Recovery Waiting for thread 1 sequence 9 (in transit)Recovery of Online Redo Log: Thread 1 Group 14 Seq 9 Reading mem 0Mem# 0: /wload/test2/app/oracle/u02/oradata/testdb1/srl14a.rdoMem# 1: /wload/test2/app/oracle/u03/oradata/testdb1/srl14b.rdoFlashback Dataguard DatabaseSteps to flashback the primary and stanby database to a point in time.Disable dataguard configurationdgmgrlshow configuration check no errors before starting thisdisable FAST_START FAILOVER;DISABLE CONFIGURATION;Primarysrvctl stop database -d PTEST1srvctl start database -d PTEST1 -o mountSQLflashback database to timestamp to_timestamp(2011-02-17 14:00:00, YYYY-MM-DD HH24:MI:SS) or restore pointSQLalter database open resetlogs;SQLselect resetlogs_change# from v$database;srvctl start service -s TEST -d PTEST1Standbysrvctl stop database -d FTEST1srvctl start database -d FTEST1 -o mountflashback database to scn Enable dataguard configurationdgmgrlENABLE CONFIGURATION;ENABLE FAST_START FAILOVER;show configuration check no errors