On 01/18/2016 09:44 AM, Paulo Motta wrote:
broadcast_address is the address exposed for internal inter-node
communication, while rpc_address is the address that will listen to
clients.
all nodes need to talk to each other via the broadcast_address, so if
they are within the same network, you may use public or private IPs as
broadcast_address, but if there's at least one node in a different
network they all need to use the public IP, or you need to setup your
own tunnelling/vpn to make sure nodes can reach each other.
You need to setup your own firewall rules. See more about what ports
are used here:
https://docs.datastax.com/en/cassandra/2.1/cassandra/security/secureFireWall_r.html.
You may also be interested in setting up client authentication:
https://docs.datastax.com/en/cassandra/2.1/cassandra/security/security_config_native_authenticate_t.html
Thanks for links/info.
For applications, do they use the CQL native clients port(9042) or
Cassandra client port (Thrift).(9160). We will be using Python to
connect to Cassandra.
Plan to use RCP_address internally, so for this case firewall not an
issue, but would like to know for future reference. Although I think one
would ideally always want applications in the same data center as the
database.