ES-2.3.4 Cluster in AWS - Transport Client not able to connect

Elastic Stack Elasticsearch
1

I have a 3 node ES cluster setup in AWS. Cluster discovery and REST endpoints are working fine. But when i try with a TransportClient , its unable to connect to the cluster. i am getting No Node available exception. Please see the exception in below gist.

And my AWS configurations here.

cloud.aws.access_key: "XXXXXXXXXXXXXXXX"
cloud.aws.secret_key: "CCCCCCCCCCCCCCCC"
plugin.mandatory: "cloud-aws"
discovery.type: "ec2"
discovery.ec2.groups: "escluster"
discovery.ec2.host_type: "private_ip"
discovery.ec2.ping_timeout: "30s"
discovery.ec2.availability_zones: "us-east-1a"
cloud.aws.region: "us-east"
discovery.zen.ping.multicast.enabled: false
discovery.zen.ping.unicast.enabled: true
discovery.zen.ping.unicast.hosts: ["PRIVATE_IP","PRIVATE_IP","PRIVATE_IP"]
network.host: PRIVATE_IP
network.bind_host: 0

It would be great , if any one can point me any issues with above configuration. I was using above config for my previous cluster ES 1.7 and Transport Client was also working fine then. Now I am migrating to ES 2.3.4. Please suggest and give me some pointers.

Thanks

2

My guess is that you changed the cluster name on some machines?

[Hobgoblin] node {#transport#-1}{172.30.1.168}{172.30.1.168:9300} not part of the cluster Cluster [elasticsearch]
3

Thank you @dadoonet . I confirmed that , all nodes are in same cluster. But from the line which you mentioned , there is a "elasticsearch" default cluster coming up. So that's a good hint. Let me check my Transport Client code and need to check whether cluster.name is not taking from property file or not.
Thanks

4

@dadoonet My bad, cluster.name was not set properly in TransportClient. Its working fine now. Thank you for the pointer.