In my Dataguard configuration on a 12c environment, when I perform a switchover from primary (
foo) to standby (
bar) and then try to sqlplus in with
user@foo, I receive the "
ORA-011033: ORACLE initialization or shutdown in progress" error instead of being able to connect to the new primary(bar).
Now, I know that this is because the connection is still pointing to the
foo address instead of the
bar address, and according to many articles, you need to create a new "virtual" service to handle the different connections, as well as a trigger to stop/start this service after a rolechange.
I did this and with the
service_names parameter set to this new "virtual" service, a switchover will work and can also switch back but when trying to sqlplus with
user@foo it throws
ORA-011033. The trigger is useless and the service continues to run on both foo and bar because it seems that having that parameter set overrides whatever the trigger tries to do.
Now if I don't define that parameter, my switchover works fine, and the triggers do their job, and when I
sqlplus user@foo, it will correctly redirect me to the
bar address. However, when I try to switchover back to
foo, I receive the following in the DG alert log:
Redo transport problem detected: redo transport to database foo has the following error:
ORA-16047: DGID mismatch between destination setting and target database 08/05/2016 16:01:02 Initialization of connection failed. Expected destination db_unique_name is foo. Database actually reached is bar Failed to send message to site foo. Error code is ORA-16642. Data Guard Broker Status Summary:
Type Name Severity Status Configuration foo_dg Warning ORA-16607 Primary Database bar Error ORA-16778 Physical Standby Database foo Error ORA-16642
log_archive_dest_2='service="bar"','ASYNC NOAFFIRM delay=0 optional compression=disable max_failure=0 max_connections=1 reopen=300 db_unique_name="bar" net_timeout=30','valid_for=(online_logfile,all_roles)' fal_client='foo' fal_server='bar'
log_archive_dest_2='service="foo", ASYNC NOAFFIRM delay=0 optional compression=disable max_failure=0 max_connections=1 reopen=300 db_unique_name="foo" valid_for=(online_logfile,all_roles)' fal_client='bar' fal_server='foo'
Entries in LDAP: foo:
I'm not sure what I'm missing or misunderstanding. Can anyone shed any light as to what I might be doing wrong? Can provide more info if needed.