Use nodetool to gather information about the cluster:

nodetool status

Take a note of the datacenter names and how many nodes are in each datacenter. Since the cluster has only two nodes, we can have at most two replicas. In a real-life production cluster, you can usually expect to have 3 or more nodes per datacenter and replication factors of 3 or higher.

To start the CQL shell and connect to a node in the cluster, either specify a node IP address, i.e., cqlsh 172.xxx.xxx.xxx.

Here are the IP addresses for the nodes in this scenario:

Node 1 - [[HOST1_IP]]
Node 2 - [[HOST2_IP]]
Node 3 - [[HOST3_IP]]

cqlsh [[HOST1_IP]]

Create a keyspace with name simple_keyspace_1 that uses SimpleStrategy and a replication factor of 1:

CREATE KEYSPACE simple_keyspace_1
WITH replication = {'class': 'SimpleStrategy', 
                    'replication_factor': 1};

The replication_factor option specifies a replication factor for the entire cluster. That means that SimpleStrategy does not respect datacenter layouts and, therefore, is not a good choice for production.

Create a keyspace with name simple_keyspace_2 that uses SimpleStrategy and a replication factor of 2. Use tab completion in cqlsh to your advantage.

Create a keyspace with name production_keyspace_1 that uses NetworkTopologyStrategy and a replication factor of 1 for DC-West:

CREATE KEYSPACE production_keyspace_1
WITH replication = {'class': 'NetworkTopologyStrategy', 
                    'DC-West': 1};

With NetworkTopologyStrategy, a replication factor is specified for each datacenter separately. In the above example, the DC-West datacenter will have a single copy of data. Not replicating data to DC-East could be a valid use case but usually there will be replicas in each datacenter.

Create a keyspace with name production_keyspace_2 that uses NetworkTopologyStrategy and 1 replica in each datacenter:

Having 3 replicas per datacenter is a good starting point for relatively small clusters. As the number of nodes in a datacenter becomes larger, a higher replication factor may become a better choice.

The number of replicas can affect consistency, availability, latency and throughput. Increasing a replication factor improves availability as it becomes possible to tolerate more replica failures. Also, a larger set of replicas can serve more concurrent requests and result in better response times.

Try the following five CQL shell commands and CQL statements that are applicable to keyspaces.

1. List the names of all keyspaces in the cluster:

   DESCRIBE KEYSPACES;
   

2. Output all CQL statements that can be used to recreate the given keyspace and all the schema objects that belong to it:

   DESCRIBE KEYSPACE production_keyspace_2;
   

3. Alter properties of the given keyspace:

   ALTER KEYSPACE production_keyspace_2
   WITH replication = 
     {'class': 'NetworkTopologyStrategy',
      'DC-West': 3, 'DC-East': 5};
   

4. Set the given keyspace as the current working keyspace:

   USE production_keyspace_2;
   

5. Remove the given keyspace and all the objects that belong to it:

   DROP KEYSPACE production_keyspace_1;
   

Here is a short quiz for you.