EC2 Aws 上的 Elasticsearch 集群无法加入集群

Question

我在一个 ec2 实例上安装了 oracle-jdk8 和 elasticsearch，并从中创建了一个 ami。在 ec2 机器的下一个副本中，我刚刚更改了 elasticsearch.yml 中的节点名称 但是，如果单独运行，两个节点都在运行。[注意节点 ID 显示为相同] 但是如果同时运行，稍后启动的节点将失败，并在日志中显示以下内容：

[2018-08-07T16:35:06,260][INFO][o.e.d.z.ZenDiscovery][node-1] 未能向主服务器发送加入请求 [{node-2}{uQHBhDuxTeWOgmZHsuaZmA}{akWOcJ47SZKpR_EpA2lpyg}{10.127.114.212}{10.127.114.212:9300}{aws_availability_zone=us-east-1c，ml.machine_memory=66718932992，ml.max_open_jobs=20， xpack.installed=true, ml.enabled=true}]，原因 [RemoteTransportException[[node-2][10.127.114.212:9300][internal:discovery/zen/join]]； 嵌套：IllegalArgumentException [无法添加节点 {node-1}{uQHBhDuxTeWOgmZHsuaZmA}{Ba1r1GoMSZOMeIWVKtPD2Q}{10.127.114.194}{10.127.114.194:9300}{aws_availability_zone=us-east-1c，ml.machine_memory=66716696576，ml.max_open_jobs=20， xpack.installed=true, ml.enabled=true}，找到现有节点 {node-2}{uQHBhDuxTeWOgmZHsuaZmA}{akWOcJ47SZKpR_EpA2lpyg}{10.127.114.212}{10.127.114.212:9300}{aws_availability_zone=us-east-1c，ml.machine_memory=66718932992，xpack.installed=true， ml.max_open_jobs=20, ml.enabled=true} 具有相同的 id 但属于 不同节点实例];

我的 elasticsearch.yml：

cluster.name: elk
node.name: node-1

path.data: /var/lib/elasticsearch
path.logs: /var/log/elasticsearch
bootstrap.memory_lock: true

network.host: 0.0.0.0
network.publish_host: _ec2:privateIp_
transport.publish_host: _ec2:privateIp_

discovery.zen.hosts_provider: ec2
discovery.ec2.tag.ElasticSearch: elk-tag

cloud.node.auto_attributes: true
cluster.routing.allocation.awareness.attributes: aws_availability_zone

来自_nodes 端点的输出：

//----Output when node-1 is run individually/at first----
{
  "_nodes" : {
    "total" : 1,
    "successful" : 1,
    "failed" : 0
  },
  "cluster_name" : "elk",
  "nodes" : {
    "uQHBhDuxTeWOgmZHsuaZmA" : {
      "name" : "node-1",
      "transport_address" : "10.127.114.194:9300",
      "host" : "10.127.114.194",
      "ip" : "10.127.114.194",
      "version" : "6.3.2",
      "build_flavor" : "default",
      "build_type" : "rpm",
      "build_hash" : "053779d",
      "roles" : [
        "master",
        "data",
        "ingest"
      ],
      "attributes" : {
        "aws_availability_zone" : "us-east-1c",
        "ml.machine_memory" : "66716696576",
        "xpack.installed" : "true",
        "ml.max_open_jobs" : "20",
        "ml.enabled" : "true"
      },
      "process" : {
        "refresh_interval_in_millis" : 1000,
        "id" : 3110,
        "mlockall" : true
      }
    }
  }
}
//----Output when node-2 is run individually/at first----
{
  "_nodes" : {
    "total" : 1,
    "successful" : 1,
    "failed" : 0
  },
  "cluster_name" : "elk",
  "nodes" : {
    "uQHBhDuxTeWOgmZHsuaZmA" : {
      "name" : "node-2",
      "transport_address" : "10.127.114.212:9300",
      "host" : "10.127.114.212",
      "ip" : "10.127.114.212",
      "version" : "6.3.2",
      "build_flavor" : "default",
      "build_type" : "rpm",
      "build_hash" : "053779d",
      "roles" : [
        "master",
        "data",
        "ingest"
      ],
      "attributes" : {
        "aws_availability_zone" : "us-east-1c",
        "ml.machine_memory" : "66718932992",
        "xpack.installed" : "true",
        "ml.max_open_jobs" : "20",
        "ml.enabled" : "true"
      },
      "process" : {
        "refresh_interval_in_millis" : 1000,
        "id" : 4869,
        "mlockall" : true
      }
    }
  }
}

Answer 1

通过在每个实例中删除 rm -rf /var/lib/elasticsearch/nodes/ 并重新启动 elasticsearch 来解决此问题。