Database Reference
In-Depth Information
be done using
nodetool
. You need to run
nodetool move
NEW_INITIAL_TOKEN
on each node.
•
Cleanup
: Cassandra does not really move the data from one machine to another;
it copies the data instead. This leaves nodes with unused old data. To clean this
data, execute
nodetool cleanup
on each node.
There is a demonstration of the addition of a node into a three-node cluster, that is, the ex-
pansion of a three-node cluster into a four-node cluster in the following section.
Adding new nodes in vnode-enabled clusters
It is simple to add new node to vnode-enabled clusters because Cassandra takes care of
distributing data (almost) evenly across the nodes without any manual intervention.
Use
nodetool -h HOSTNAME status
to see the cluster status:
$ bin/nodetool status
Note: Ownership information does not include topology; for
complete information, specify a keyspace
Datacenter: us-east
===================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
-- Address Load Tokens Owns Host
ID Rack
UN 10.10.21.228 1.88 GB 256 35.2% d3de1... 1a
UN 10.10.21.206 1.86 GB 256 33.1% 7dced... 1a
UN 10.10.21.169 1.55 GB 256 31.7% 55ac5... 1a
So, our cluster is more or less balanced. Let's start a new node, and check the status:
$ bin/nodetool status
Note: Ownership information does not include topology; for
complete information, specify a keyspace
Datacenter: us-east
===================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
-- Address Load Tokens Owns Host
ID Rack
