question

@amit82ocp_172818 In your application, you would typically configure the driver with the contact points (seed nodes, if you like) and explicitly specify the local datacenter. This will cause the driver to compute a query plan which only contains nodes from the local datacenter.

To illustrate with an example, here is how contact points from the datacenter DC1 are configured in the Java driver v4.6:

CqlSession session = CqlSession.builder()
    .addContactPoint(new InetSocketAddress("10.1.2.3", 9042))
    .addContactPoint(new InetSocketAddress("10.1.2.4", 9042))
    .withLocalDatacenter("DC1")
    .build();

In this example, we are telling the driver that nodes 10.1.2.3 and 10.1.2.4 belong to the DC1 datacenter. When the driver computes a query plan, it will only ever include nodes in DC1 in the list of nodes to connect to.

Alternatively, you can also provide the contacts points to the driver in a configuration file like this:

datastax-java-driver {
  basic {
    contact-points = [ "10.1.2.3:9042", "10.1.2.4:9042" ]
    load-balancing-policy.local-datacenter = DC1
  }
}

For more information, see the Core driver and Load balancing pages of the Java Driver. For information about other drivers, see the Driver matrix page. Cheers!