Joining new Cassandra nodes to an existing Cassandra cluster

The third step to adding data centers to an existing setup is to join the newly created Cassandra nodes to the existing Cassandra cluster.

Before you begin

Before you join the new Cassandra nodes, ensure that the following steps are completed:

  1. Setting up a shared file system
  2. Installing Cassandra on the new data center

Joining the Cassandra nodes

Joining the Cassandra nodes is a manual process that includes the following steps:

  1. Alter namespaces through cqlsh.
  2. Modify the Cassandra configuration files.
  3. Run the nodetool utility on every Cassandra node.

Refer to the Datastax documentation for detailed steps for joining the Cassandra nodes: Adding a data center to a cluster.

Note: Until the data in the new data center nodes is synchronized with the existing data centers and the new data center joins the cluster, queries that are made with high consistency, such as QUORUM, EACH_QUORUM, might fail on the existing data centers.

What to do next

Ensure that the new data center is joined to the cluster and that all data centers are replicating data with each other. The nodetool rebuild must have run on all new Cassandra nodes.