datastax - Cassandra NoHostAvailableException; How to set up Native protocol Cassandra cluster? -

i know question has been asked around internet lot, including stackoverflow, far none of solutions have worked me. i'm thinking might weird error .yaml file

1) know not firewall issue, i've tested connection between 2 nodes

2) i've increased timeouts

3) i've set rpc address 0.0.0.0 , public ip

4) i've set broadcast address public ip

couple irregularities

1) noticed on first cluster (which works) output cqlsh:

connected cas1 @ 127.0.0.1:9042. [cqlsh 5.0.1 | cassandra 2.1.3 | cql spec 3.2.0 | native protocol v3]

now, on new cluster is:

connected test cluster @ localhost:9160. [cqlsh 4.1.1 | cassandra 2.0.16 | cql spec 3.1.1 | thrift protocol 19.39.0

is normal? set process wasn't different.

2) when put in listen address in cas1 cluster, ip address white. when put in listen address in test cluster, ip address becomes red 1 of "." white. i'm not expert on stuff, know isn't normal. i've tried c/p , hasn't helped

3) noticed test cluster didn't have "broadcast_rpc_address" setting cas1 cluster did, c/p on (obv changed address) didn't work well

so both of settings same in 2 cassandra clusters:

listen address: private.ip

start_native_transport: true

native_transport_port:9042

start_rpc:true

rpc_address: public.ip //also tried 0.0.0.0

rpc_port:9160

broadcast_rpc_address:private ip //also tried public

4) in meantime, i'm going try re install cassandra, far in details haven't seen anyway on how can configure if native or thrift protocol, please let me know if there way can specify this.

thanks help!

all modern versions speak both thrift (9160) , native (9042).

native used cqlsh in 2.1+, cqlsh in 2.0 prefers thrift.

you've somehow mixed versions - because did apt/yum install without paying attention versions. it's worth noting 0.0.0.0 ips shortcut doesn't work - right ips in there, , make sure it's binding right ips.