Capture and Apply Processes in a RAC
Instance
--Link Name matches the RAC DB name, but
use the TNS entry that
-- points to the Instance : DNYDBA2A
CREATE DATABASE LINK DNYDBA20.world CONNECT
TO strmadm IDENTIFIED BY
strmadm USING ' DNYDBA2A.world'
/
When the DNYDBA2A fails, the queue moves
over to the DNYDBA2B instance automatically. In that case, at
the DNYTST10 site, the DNYDBA20.world link should be dropped and
recreated with the same name but with the TNS entry that points
to the DNYDBA2B instance. Recreate the database link as shown
below:
CREATE DATABASE LINK DNYDBA20.world CONNECT
TO strmadm IDENTIFIED BY strmadm USING 'DNYDBA2B.world'
/
Since the database link name is not
changed, the propagation process attributes, if any at the
DNYTST10 site, remain the same.
At the same time, for the RAC database
DNYDBA20 to propagate to the DNYTST10, there is no need for any
change in the configuration. Once the queue moves over to the
DNYDBA2B instance, the propagation also moves over and resumes
the replication to the destination of DNYTST10.
The following output shows the failover of
queue buffers. It shows that the RAC database has two instances
and the NY1_QUEUE_TABLE is currently on Instance-1, which is
DNYDNA2A.
