在 Windows 10 台式机上启动了三个 Elasticsearch 的节点，版本是 7.10.1，将配置文件分别改为如下内容：

# master node
cluster.name: es-study-cluster
node.name: master
network.host: 127.0.0.1
http.port: 9200
# slave-1
cluster.name: es-study-cluster
node.name: slave-1
network.host: 127.0.0.1
http.port: 9201
# slave-2
cluster.name: es-study-cluster
node.name: slave-2
network.host: 127.0.0.1
http.port: 9202

按照在网络上看到的教程，三个节点应该自动加入一个集群，形成一主两备的局面，结果……

2号“奴隶”节点自立山头了。

在浏览器分别查看 Elasticsearch 的状态如下：

# master
{
  "name" : "master",
  "cluster_name" : "es-study-cluster",
  "cluster_uuid" : "ras0xD8ARh-WGaJ4VwUY1Q",
  "version" : {
    "number" : "7.10.1",
    "build_flavor" : "default",
    "build_type" : "zip",
    "build_hash" : "1c34507e66d7db1211f66f3513706fdf548736aa",
    "build_date" : "2020-12-05T01:00:33.671820Z",
    "build_snapshot" : false,
    "lucene_version" : "8.7.0",
    "minimum_wire_compatibility_version" : "6.8.0",
    "minimum_index_compatibility_version" : "6.0.0-beta1"
  },
  "tagline" : "You Know, for Search"
}

# slave-1
{
  "name" : "slave-1",
  "cluster_name" : "es-study-cluster",
  "cluster_uuid" : "ras0xD8ARh-WGaJ4VwUY1Q",
  "version" : {
    "number" : "7.10.1",
    "build_flavor" : "default",
    "build_type" : "zip",
    "build_hash" : "1c34507e66d7db1211f66f3513706fdf548736aa",
    "build_date" : "2020-12-05T01:00:33.671820Z",
    "build_snapshot" : false,
    "lucene_version" : "8.7.0",
    "minimum_wire_compatibility_version" : "6.8.0",
    "minimum_index_compatibility_version" : "6.0.0-beta1"
  },
  "tagline" : "You Know, for Search"
}

# slave-2
{
  "name" : "slave-2",
  "cluster_name" : "es-study-cluster",
  "cluster_uuid" : "PtZ0JU5bQBG_OBXh3tMyVQ",
  "version" : {
    "number" : "7.10.1",
    "build_flavor" : "default",
    "build_type" : "zip",
    "build_hash" : "1c34507e66d7db1211f66f3513706fdf548736aa",
    "build_date" : "2020-12-05T01:00:33.671820Z",
    "build_snapshot" : false,
    "lucene_version" : "8.7.0",
    "minimum_wire_compatibility_version" : "6.8.0",
    "minimum_index_compatibility_version" : "6.0.0-beta1"
  },
  "tagline" : "You Know, for Search"
}

可以注意到 slave-2 的集群名称 cluster_name 没有问题，但是 cluster_uuid 与 master 和 slave-1 不一样。

作为一个 Elasticsearch 的新人，我有点手足无措。

为了证明是 slave-2 的个人情绪问题，所以我又启动了一个 slave-3，然后发现 slave-3 很顺利的就融入了集体。

{
  "name" : "slave-3",
  "cluster_name" : "es-study-cluster",
  "cluster_uuid" : "ras0xD8ARh-WGaJ4VwUY1Q",
  "version" : {
    "number" : "7.10.1",
    "build_flavor" : "default",
    "build_type" : "zip",
    "build_hash" : "1c34507e66d7db1211f66f3513706fdf548736aa",
    "build_date" : "2020-12-05T01:00:33.671820Z",
    "build_snapshot" : false,
    "lucene_version" : "8.7.0",
    "minimum_wire_compatibility_version" : "6.8.0",
    "minimum_index_compatibility_version" : "6.0.0-beta1"
  },
  "tagline" : "You Know, for Search"
}

然后去看了一下日志，发现了一些警告信息

[2021-01-12T12:34:05,308][WARN ][o.e.g.DanglingIndicesState] [MS-FZYRJUKHIOOR] gateway.auto_import_dangling_indices is disabled, dangling indices will not be automatically detected or imported and must be managed manually
...
[2021-01-12T12:34:07,103][WARN ][o.e.d.HandshakingTransportAddressConnector] [MS-FZYRJUKHIOOR] handshake failed for [connectToRemoteMasterNode[127.0.0.1:9301]]
java.lang.IllegalStateException: handshake with [{127.0.0.1:9301}{XLkiFTcWRGabfDE7s5nzrg}{127.0.0.1}{127.0.0.1:9301}] failed: remote cluster name [es-study-cluster] does not match local cluster name [elasticsearch]
...
[2021-01-13T17:21:30,921][WARN ][o.e.h.AbstractHttpServerTransport] [slave-1] caught exception while handling client http traffic, closing connection Netty4HttpChannel{localAddress=/127.0.0.1:9201, remoteAddress=/127.0.0.1:49243}
java.io.IOException: 你的主机中的软件中止了一个已建立的连接。

从日志里面看，似乎是第二个节点多次试图加入 cluster，但是不知道什么原因被拒绝了。然后放狗去查，在万能的 stackoverflow 上找到了一个，在 slave-2 的配置文件里面增加了一行

discovery.zen.ping.unicast.hosts: ["127.0.0.1:9300", "127.0.0.1:9301"]

然后……

slave-2 就高高兴兴的加入小团伙了。

另外一个让我不理解的事情，就是我的启动顺序是 master，slave-1，slave-2，可是从 elasticsearch-head 里面来看，似乎 slave-1 抢班夺权，成了主节点，不知道是为什么。

作为小白，还有一个问题，就是为什么题图上的 shard 上面有编号，而我的都是 0。

是说我从零开始学 Elasticsearch 么？