Formula for stackd.io that provisions a cluster with Cassandra 1.2.x (DataStax Community Edition)
Seed nodes in Cassandra are the gatekeepers of the cluster topology and network addresses for all the other nodes in the cluster. When nodes are brought online they must know at least one seed node so they can figure out where all the other nodes are and where in the cluster they themselves fit. Most of this is handled by Cassandra, but we explicitly separate regular and seed nodes so we can target them individually. This makes it much simpler to write the SLS and obtain the seed node hostnames.
NOTE: The order here should be lower than both regular nodes and OpsCenter nodes to ensure that seeds are available before regular nodes and OpsCenter attempts to connect.
This component will provision a machine as a regular Cassandra node. They will be configured to connect to Seed Nodes, so their order must be higher than any seed nodes' order.
[Optional]
DataStax's OpsCenter provides a great dashboard to help you manage your Cassandra cluster. Simply designate one of your nodes to use this component (or use an entirely separate machine), set its order to be higher than seed and regular nodes, and you'll have a web application to help you with all of your Cassandra needs.
PLEASE NOTE: Only one of these should be set for any given cluster and the order for OpsCenter should be higher than any other component because it depends on the complete Cassandra cluster to be up and running first.