Will you please let us know the consume code for reading two different topic?

If you writing to primary cluster at normal time, why to keep reading from both the topic.

Is there any unidirectional replication happening between primary and secondary server.