DataGuard Switchover and Back

[Former Member]

Dear All

I have configured successfully DataGuard and Broker. Now I am looking successfully switchover and back.

Could you please provide the stpes guide  line or document

Regards

Vimal

[thiago_cavalheiro]

Hello Vimal,

I recommend you to check Oracle documentation which is thorough on the topic:

http://docs.oracle.com/cd/E11882_01/server.112/e17023/sofo.htm

Hope this helps.

Cheers,

Thiago

[Former Member]

Hello Thiago

Thanks for reply..

I have gone thru the document. I would like to have the steps by steps process in form of command for switchover and switchback.

Thanks in advance

Regards

Vimal Pathak

[ACE-SAP]

Hello

All you need to know / do is in the documentation quoted by Thiago...

If you are running db_node1 as primary you will then need to issue the following command form broker console in order to switch to standby node.

DGMGRL> switchover to db_node2;

Then to switch back just type

DGMGRL> switchover to db_node1;

Watch out data guard can look simple but if misused it could be a very efficient tool to scratch your DB.

BTW, make sure your DB are running in forced logging mode.

Best regards

[Former Member]

Dear Yves

Thanks for reply

I have tried the switchover  as below:

DGMGRL> switchover to RQ9_DRS;

And this is working successfully. But when I tried to switch back with below command :

DGMGRL> switchover to RQ9_PRD 

Error : "rq9_prd" was not found;

While in the configuration, it is showing as below:

DGMGRL>
enable configuration
Enabled.

DGMGRL> show configuration

Configuration
- standby_config

Protection Mode: MaxPerformance
  Databases:
    rq9_drs - Primary database
    RQ9_PRD - Physical standby database

Fast-Start
Failover: DISABLED

Configuration
Status:
SUCCESS

DGMGRL>
switchover to RQ9_PRD
Site "rq9_prd" was not found
DGMGRL>

PLease suggest.

Regards

Vimal Pathak

[ACE-SAP]

Hello

Please provide full info from broker

=> show configuration verbose

=> show database verbose rq9_drs

=> show database verbose RQ9_PRD

Could also be interesting to have the archive dest configuration on both nodes

=> show parameter archive

Best regards

[Former Member]

Hi Yves

I stuck in a last issue.

I have started to do the failover scenarion.... I followed below steps:

1. Shutdown the primary production server

2.  failover to standby server

3. Now the primary is in physical standby mode and standby database is in primary role

4. Tries to reinstate the database and getting the error as below

ORA-16795: the standby database need to be re-created

Failed

configuration detail can not be determinded by DGMGRL.

Please note that in between I have tried the "alter system set dg_broker_start=false" and then  "alter system set dg_broker_start=true"

Please suggest what need to be done.

Regards

Vimal

[stefan_koehler]

Hi Vimal,

this is not an issue - this is works as designed. Data Guard is nothing to implement by reading "a snapshot to snapshot" guide. It seems like you got no training on this and forgot to enable "flashback database".

I already discussed it in this SCN thread here.

Regards

Stefan

[Former Member]

Hi Stefan

You are right. I forget to  enable the "flashback database". what is work around. My standby database is now working fine as primary role and old primary database which is in standby mode is giving below error:

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

Managed
Standby Recovery starting Real Time Apply

Datafile 1 is on orphaned branch

          File status = 8196

        Abs fuzzy SCN = 0

Hot backup fuzzy SCN = 145008338

MRP0: Detected orphaned datafiles!

Recovery will possibly be retried after flashback...

Errors in file
E:\ORACLE\RQ9\SAPTRACE\diag\rdbms\rq9_prd\rq9\trace\rq9_pr00_4076.trc:

ORA-19909: datafile 1 belongs to an orphan incarnation

ORA-01110: data file 1: 'E:\ORACLE\RQ9\SAPDATA1\SYSTEM_1\SYSTEM.DATA1'

Managed Standby Recovery not using Real Time Apply

Slave exiting with ORA-19909 exception

Errors in file E:\ORACLE\RQ9\SAPTRACE\diag\rdbms\rq9_prd\rq9\trace\rq9_pr00_4076.trc:

ORA-19909: datafile 1 belongs to an orphan incarnation

ORA-01110: data file 1: 'E:\ORACLE\RQ9\SAPDATA1\SYSTEM_1\SYSTEM.DATA1'

Completed: ALTER DATABASE RECOVER MANAGED STANDBY DATABASE  THROUGH ALL
SWITCHOVER DISCONNECT  USING CURRENT LOGFILE

Recovery Slave PR00 previously exited with exception 19909

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

I have check the standby server. the "flashback is on" and when I am trying to reinstate the database at primary get below error

  DGMGRL>
connect sys@RQ9_PRD
Password:
Connected.

DGMGRL> show configuration

  Configuration
- standby_config

Protection Mode: MaxPerformance
  Databases:
    rq9_drs - Primary database
    RQ9_PRD - Physical standby database
      Error: ORA-16766: Redo Apply is stopped

  Fast-Start
Failover: DISABLED

  Configuration
Status:
ERROR

DGMGRL>
reinstate database "RQ9_PRD"
Reinstating database "RQ9_PRD", please wait...
Error: ORA-16653: failed to reinstate database

  Failed.
Reinstatement of database "RQ9_PRD" failed
DGMGRL>

Pl suggest the solution

Regards

Vimal pathak

[stefan_koehler]

Hi Vimal,

there is no work around as it is not a bug. You need to perform a full restore and recovery of your "new standby database" RQ9_PRD. That's it - by the way it is also mentioned in the SCN thread from above.

Regards

Stefan