Few days ago i got the activity to create the 2 NODE RAC DR on EXADATA for 2 NODE RAC Production.

Here I am listing few issues that made this activity perfect case for Data Guard beginners

Environment:
Production Database Name: Finance
db_unique_name : finance

Standby Database Name : Finstandby
db_unique_name : finstandby

Production OS Node: agoracledba1 agoracledba2
Standby OS Node: agoracledba3 agoracledba4

The Standby is created with below script:
startup with pfile in nomount stage

Connect to database and start in nomunt

Now create the standby database

Run the RMAN script for creating standby

This script will restore the standby from the live production database over the network.


Everything was fine till the recovery of the database is not started, recovery is cancelled and error is reported in alert log .
Looking in the Alert Log one will find unusual error as listed below.



Now connect to standby database and check the name of datafiles.

From Prodcution database exact names of the datafile can be find and will be used to restore the datafiles with original names.


Same Datafiles need to be restored with new datafile name by using “SET NEW NAME” parameter as no datafile is available physically on ASM DISKs.

Rerun the RMAN script

this will restore all the missing datafiles and can be check at the standby database by running the below query.

For Consistency in the database I am suggesting to restore the whole database using rman by restoring CONTROLFILE first and then FULL DATABASE RESTORE to avoid any other unusual error.

At standby, rerun the above query again

Now, check the errors datafile again with below query.

See all the required datafiles are restored and the standby database is consistent.


While restoring db or required datafiles , restoration progress can be checked by the following query:


Error No 2

The ORA-19909 error occurs when you have made an error choosing a restore file that is part of the current (or prior) incarnation of the database. The oerr utility show this for the ORA-19909 error:

ORA-19909: datafile %s belongs to an orphan incarnation

Cause: Either the specified datafile was restored from a backup that was taken during a period of time that has already been discarded by a resetlogs operation, or Oracle cannot identify which database incarnation the file belongs to. The alert log contains more information.

Action: Restore a backup of this file that belongs to either the current or a prior incarnation of the database. If you are using RMAN to restore, RMAN will automatically select a correct backup.

You can run this command to see the correct incarnation:

To verify that you have the correct incarnation, you can display the checkpoint_change# and checkpoint_time columns in v$datafile and v$datafile_header and compare these values with those found in v$database.

One solution is to reset the standby database’s incarnation to match the primary database incarnation:

This blog is completed.

Thank you for giving your valuable time to read the above information.

If you want to be updated with all our articles send us the Invitation or Follow us:

Telegram Channel: https://t.me/helporacle

For any assistance: http://www.facebook.com/er.arungupta
Skant Gupta’s LinkedIn: www.linkedin.com/in/skantali/

Joel Perez’s LinkedIn: Joel Perez’s Profile

LinkedIn Group: Oracle Cloud DBAAS

Facebook Page: OracleHelp

Comments

  1. SriHarsha

    You can use db_create_file_dest and db_create_online_log_dest*. No need to use file_convert for both data and log files.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.