es集群配置

2019-02-13

服务器部署说明

Linux

安装wget yum install wget -y
修改内核最大map数量并生效 echo 'vm.max_map_count = 262144' >> /etc/sysctl.conf && sysctl -p
新增用户es useradd es -d /opt/es
设置文件打开数量编辑 vi /etc/security/limits.conf 增加 es - nofile 65536

elasticsearch 集群部署

查看各个节点ip,确保能相互ping通: 如:
```
10.20.250.80
10.20.250.81
10.20.250.82
```

各个节点切换用户,下载elasticsearch,解压

su - es
wget https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-6.3.1.tar.gz
tar -zxv -f elasticsearch-6.3.1.tar.gz
ln -s elasticsearch-6.3.1 elasticsearch

编辑节点1配置文件 elasticsearch/config/elasticsearch.yml

    cluster.name: erp-dfs
    node.name: ${HOSTNAME}
    #path.data: /esdata/data
    #path.logs: /esdata/logs
    network.host: 10.20.250.80
    http.port: 9200
    transport.tcp.port: 9300
    bootstrap.system_call_filter: false
    xpack.monitoring.collection.enabled: true
    discovery.zen.ping.unicast.hosts: ["10.20.250.81", "10.20.250.82"]

编辑节点2配置文件 elasticsearch/config/elasticsearch.yml

    cluster.name: erp-dfs
    node.name: ${HOSTNAME}
    #path.data: /esdata/data
    #path.logs: /esdata/logs
    network.host: 10.20.250.81
    http.port: 9200
    transport.tcp.port: 9300
    bootstrap.system_call_filter: false
    xpack.monitoring.collection.enabled: true
    discovery.zen.ping.unicast.hosts: ["10.20.250.80", "10.20.250.82"]

编辑节点3配置文件 elasticsearch/config/elasticsearch.yml

    cluster.name: erp-dfs
    node.name: ${HOSTNAME}
    #path.data: /esdata/data
    #path.logs: /esdata/logs
    network.host: 10.20.250.82
    http.port: 9200
    transport.tcp.port: 9300
    bootstrap.system_call_filter: false
    xpack.monitoring.collection.enabled: true
    discovery.zen.ping.unicast.hosts: ["10.20.250.80", "10.20.250.81"]

启动各个节点

    bin/elasticsearch

验证各个节点是否正常运行查看启动信息中是否包含以下信息:

    [2018-07-31T12:39:18,667][INFO ][o.e.c.s.ClusterApplierService] [sysopt-node3.novalocal] added {{sysopt-node1.novalocal}{N8IhoG11Tfia8Y52klYrqg}{bqqmJB2iScaV1f2rD9qCyw}{10.20.250.80}{10.20.250.80:9300}{ml.machine_memory=8202297344, ml.max_open_jobs=20, xpack.installed=true, ml.enabled=true},}, reason: apply cluster state (from master [master {sysopt-node3.novalocal}{adTO7mKPT32UAq9aZxY_gg}{0C_DuqnPReSxEZ6i2hO5Hw}{10.20.250.82}{10.20.250.82:9300}{ml.machine_memory=8202297344, xpack.installed=true, ml.max_open_jobs=20, ml.enabled=true} committed version [15] source [zen-disco-node-join]])
    [2018-07-31T12:40:59,787][INFO ][o.e.c.s.ClusterApplierService] [sysopt-node3.novalocal] added {{sysopt-node2.novalocal}{-CHWSrLjQTOZqT3q75GVIA}{r93bo_kQTwOZxNZCekDsWA}{10.20.250.81}{10.20.250.81:9300}{ml.machine_memory=8202297344, ml.max_open_jobs=20, xpack.installed=true, ml.enabled=true},}, reason: apply cluster state (from master [master {sysopt-node3.novalocal}{adTO7mKPT32UAq9aZxY_gg}{0C_DuqnPReSxEZ6i2hO5Hw}{10.20.250.82}{10.20.250.82:9300}{ml.machine_memory=8202297344, xpack.installed=true, ml.max_open_jobs=20, ml.enabled=true} committed version [16] source [zen-disco-node-join]])

    netstat -tunlp |grep 9200
    netstat -tunlp |grep 9300

结果:

    tcp6       0      0 10.20.250.82:9200       :::*                    LISTEN      10373/java
    tcp6       0      0 10.20.250.82:9300       :::*                    LISTEN      10373/java

```
是否有输出
curl http://10.20.250.82:9200 

{
"name" : "sysopt-node1.novalocal",
"cluster_name" : "erp-dfs",
"cluster_uuid" : "NclsYFudSHOgBxGHoap2yQ",
"version" : {
    "number" : "6.3.1",
    "build_flavor" : "default",
    "build_type" : "tar",
    "build_hash" : "eb782d0",
    "build_date" : "2018-06-29T21:59:26.107521Z",
    "build_snapshot" : false,
    "lucene_version" : "7.3.1",
    "minimum_wire_compatibility_version" : "5.6.0",
    "minimum_index_compatibility_version" : "5.0.0"
},
"tagline" : "You Know, for Search"
}


8. 启用守护进程

> 停止各个节点服务,使用下面的命令进行守护

    bin/elasticsearch -d -p pid