NoNodeAvailableException [没有已配置的节点可用]在Elasticsearch 5.1.2中

Question

On Google Compute Engine, I created 3 VMs and installed Elasticsearch 5.1.2 on them. I installed GCE Discovery Plugin for unicast discovery.

From my local web browser(Win7), I could access these Elasticsearch nodes successfully. On Google Cloud Platform, I have added the firewall rule accepting tcp:9300, tcp:9200.

Now I'd like to use Java transport client to talk to remote Elasticsearch nodes from my local java application. I'm sure cluster.name is correct.

Code and Error are as follows:

public class NativeClient {

    @SuppressWarnings({ "resource", "unchecked" })
    public static Client createTransportClient() throws UnknownHostException {

        Settings settings = Settings.builder().put("cluster.name", "elasticsearch").put("client.transport.sniff", true)
                .build();

        TransportClient client = new PreBuiltTransportClient(settings)
                .addTransportAddress(new InetSocketTransportAddress(InetAddress.getByName("104.100.100.96"), 9300));

        return client;
    }

    private final Node node = null;
    private final Client client = null;
}


public class IndicesOperations {
    private final Client client;

    public IndicesOperations(Client client) {
        this.client = client;
    }

    public boolean checkIndexExists(String name){
        IndicesExistsResponse response=client.admin().indices().prepareExists(name).execute().actionGet();
        return response.isExists();
    }

    public static void main( String[] args ) throws InterruptedException, UnknownHostException {
        Client client =NativeClient.createTransportClient();
        IndicesOperations io=new IndicesOperations(client);
        String myIndex = "test";
        if(io.checkIndexExists(myIndex))
            io.deleteIndex(myIndex);
        io.createIndex(myIndex);
        Thread.sleep(1000);
        io.closeIndex(myIndex);
        io.openIndex(myIndex);
        io.deleteIndex(myIndex);
    }
}


Exception in thread "main" NoNodeAvailableException[None of the configured nodes are available: [{#transport#-1}{lu8DzekbSWOrNEgFgXxpgQ}{104.100.100.96}{104.100.100.96:9300}]]
    at org.elasticsearch.client.transport.TransportClientNodesService.ensureNodesAreAvailable(TransportClientNodesService.java:328)
    at org.elasticsearch.client.transport.TransportClientNodesService.execute(TransportClientNodesService.java:226)
    at org.elasticsearch.client.transport.TransportProxyClient.execute(TransportProxyClient.java:59)
    at org.elasticsearch.client.transport.TransportClient.doExecute(TransportClient.java:345)
    at org.elasticsearch.client.support.AbstractClient.execute(AbstractClient.java:403)
    at org.elasticsearch.client.support.AbstractClient$IndicesAdmin.execute(AbstractClient.java:1226)
    at org.elasticsearch.action.ActionRequestBuilder.execute(ActionRequestBuilder.java:80)
    at org.elasticsearch.action.ActionRequestBuilder.execute(ActionRequestBuilder.java:54)
    at com.packtpub.IndicesOperations.checkIndexExists(IndicesOperations.java:16)
    at com.packtpub.IndicesOperations.main(IndicesOperations.java:49)

elasticsearch.yml

network.host: _gce_

cloud:
  gce:
      project_id: es-cloud
      zone: asia-east1-b
discovery:
      type: gce

Edit :

After deploying my java application on google compute engine, it can access the elasticsearch instance running on google compute engine. In this application, I just modified InetAddress.getByName("10.140.0.2") . When deploying on my local machine, I used the external ip of that VM.

What else do I have to modify to run it on my local machine?

Answer 1

I appended the external ip of my VM to network.publish_host property in elasticsearch.yml, then I could access the elasticsearch running on remote VM:

network.host: _gce_
network.publish_host: 104.100.100.96

cloud:
  gce:
      project_id: es-cloud
      zone: asia-east1-b
discovery:
      type: gce

I don't understand exactly, but fortunately it works.