Automate Install Elasticsearch Cluster with Ansible Playbook

集群概述

脚本会在此用户下安装三节点集群，因为 7.X 版本后会自带 JDK，所以我们不需要提前安装 Java 环境。

安装计划

服务器规划

IP地址SSH 端口SSH 用户名SSH 密码ROOT 密码OS
10.1.207.18022022elasticsearch123456root123CentOS Linux release 7.9.2009
10.1.207.18122022elasticsearch123456root123CentOS Linux release 7.9.2009
10.1.207.18222022elasticsearch123456root123CentOS Linux release 7.9.2009

提示： 可以参考批量自动化创建用户

集群节点规划

IP地址Elasticsearch
10.1.207.180✓
10.1.207.181✓
10.1.207.182✓

节点安装路径

路径描述
/opt/elasticsearch程序安装路径
~/elasticsearch_uninstall.sh集群卸载脚本
~/elasticsearch.sh集群启停脚本
/data01/elasticsearch/logs日志文件
/data01/elasticsearch/data数据文件
/data01/elasticsearch/dumpDUMP 文件
/data01/elasticsearch/temp临时文件

下载安装包和 Playbook 脚本

在你的笔记本上创建 playbook 脚本存放目录

mkdir -p ~/my-docker-volume/ansible-playbook

下载 playbook 脚本

cd ~/my-docker-volume/ansible-playbook
git clone https://github.com/coolbeevip/ansible-playbook.git

下载 elasticsearch 安装包到 ~/my-docker-volume/ansible-playbook/packages 目录

wget -P ~/my-docker-volume/ansible-playbook/packages https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-7.13.3-linux-x86_64.tar.gz --no-check-certificate

配置安装脚本

您需要根据实际情况修改如下配置文件，例如 IP 地址，端口号等

main.yml

这个文件中主要定义了目标服务器的地址，登录用户名以及每个 elasticsearch 节点的名称

- hosts: 10.1.207.180
  user: elasticsearch
  ...
- hosts: 10.1.207.181
  user: elasticsearch
  ...
- hosts: 10.1.207.182
  user: elasticsearch
  ...

vars_elasticsearch.yml

这个文件中定义了安装路径、数据路径、安装包文件名、端口、内存等变量信息，你可以根据实际情况修改这些变量，这些变量在并部署时会替换到以下文件中。

操作系统 Limits

limits_hard_nproc: '65535'
limits_soft_nproc: '65535'
limits_hard_nofile: '278528'
limits_soft_nofile: '278528'
limits_hard_stack: 'unlimited'
limits_soft_stack: 'unlimited'

安装用的用户名、用户组

es_user: "elasticsearch"
es_group: "elasticsearch"

安装介质名称以及解压后的目录名

es_tar: "elasticsearch-7.13.3-linux-x86_64.tar.gz"
es_tar_unzip_dir: "elasticsearch-7.13.3"

安装路径

es_home_dir: "/opt/elasticsearch"
es_log_dir: "/data01/elasticsearch/logs"
es_data_dir: "/data01/elasticsearch/data"
es_heap_dump_path: "/data01/elasticsearch/dump"
es_temp_dir: "/data01/elasticsearch/temp"

服务配置

# 网路信息
es_network_host: "0.0.0.0"
es_http_port: 39200
es_transport_port: 39010
# 集群名称
es_cluster_name: "my-elasticsearch"
# 自动创建索引
es_action_auto_create_index: "true"
# 内存配置
es_heap_size: 8g
es_bootstrap_memory_lock: false

集群名称配置

# 集群节点名称（注意这不是主机名)
node_names:
  10.1.207.180:
    es_node_name: node-180
  10.1.207.181:
    es_node_name: node-181
  10.1.207.182:
    es_node_name: node-182

提示： 其他更多默认配置，请参考如下文件

  • ansible-playbook/elasticsearch/config/elasticsearch.yml.j2
  • ansible-playbook/elasticsearch/config/jvm.options.j2
  • ansible-playbook/elasticsearch/config/jvm.options.d/gc.options.j2

开始安装

启动 Ansible 工具连接到目标服务器，并将 ~/my-docker-volume/ansible-playbook 目录挂在到容器中。

提示： ANSIBLE_SSH_USERS，ANSIBLE_SSH_PASSS 配置成您之前在目标服务器上创建的用户名 elasticsearch 和密码 123456

提示： ANSIBLE_SU_PASSS 为 root 用户的密码

docker run --name ansible --rm -it \
  -e ANSIBLE_SSH_HOSTS=10.1.207.180,10.1.207.181,10.1.207.182 \
  -e ANSIBLE_SSH_PORTS=22022,22022,22022 \
  -e ANSIBLE_SSH_USERS=elasticsearch,elasticsearch,elasticsearch \
  -e ANSIBLE_SSH_PASSS=123456,123456,123456 \
  -e ANSIBLE_SU_PASSS=root123,root123,root123 \
  -v ~/my-docker-volume/ansible-playbook:/ansible-playbook \
  coolbeevip/ansible:2.8.11-alpine \
  /bin/bash

安装 Elasticsearch 集群

bash-5.0# ansible-playbook -C /ansible-playbook/elasticsearch/main.yml

如果你看到如下信息，说明安装完成

TASK [Install Succeed] ********************************************************************************************************************************************************************************************************
ok: [10.1.207.180] => {
    "msg": "Install Succeed!"
}

提示: 因为第一次执行脚本时，会上传 Elasticsearch 安装包到所有服务器（约 327MB），所以执行时间较长（取决于你的客户端和服务器之间的网络速度）。 你也可以在执行以上脚本前手动将安装包上传到服务器的安装路径 /opt/elasticsearch 下。在我本地环境首次安装大概耗时 5 分钟（上传安装包大概 2 分钟，安装集群大概 3 分钟）

提示: 此脚本只适合初始化安装，重复执行此命令可能会收到 Elasticsearch has been installed, please uninstall and then reinstall 提示，此时需要先要使用 ansible all -m shell -a '~/elasticsearch_uninstall.sh' 命令卸载之前的安装。

提示: 您可以使用以下脚本删除不在使用的安装包

bash-5.0# ansible all -m shell -a "rm -rf /opt/elasticsearch/elasticsearch-7.13.3-linux-x86_64.tar.gz"

验证 Elasticsearch 集群

查看目标服务器上进程, 你可以看到每个服务上有两个进程

bash-5.0# ansible all -m shell -a 'ps aux | grep [/]opt/elasticsearch | wc -l'
10.1.207.182 | CHANGED | rc=0 >>
2

10.1.207.180 | CHANGED | rc=0 >>
2

10.1.207.181 | CHANGED | rc=0 >>
2

查看目标服 elasticsearch 服务，可以看到每个节点服务都已经启动

bash-5.0# ansible all -m shell -a 'curl http://0.0.0.0:39200/?pretty'
10.1.207.182 | CHANGED | rc=0 >>
{
  "name" : "node-182",
  "cluster_name" : "nc-elasticsearch",
  "cluster_uuid" : "_na_",
  "version" : {
    "number" : "7.13.3",
    "build_flavor" : "default",
    "build_type" : "tar",
    "build_hash" : "5d21bea28db1e89ecc1f66311ebdec9dc3aa7d64",
    "build_date" : "2021-07-02T12:06:10.804015202Z",
    "build_snapshot" : false,
    "lucene_version" : "8.8.2",
    "minimum_wire_compatibility_version" : "6.8.0",
    "minimum_index_compatibility_version" : "6.0.0-beta1"
  },
  "tagline" : "You Know, for Search"
}  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   522  100   522    0     0   3279      0 --:--:-- --:--:-- --:--:--  3303

10.1.207.181 | CHANGED | rc=0 >>
{
  "name" : "node-181",
  "cluster_name" : "nc-elasticsearch",
  "cluster_uuid" : "_na_",
  "version" : {
    "number" : "7.13.3",
    "build_flavor" : "default",
    "build_type" : "tar",
    "build_hash" : "5d21bea28db1e89ecc1f66311ebdec9dc3aa7d64",
    "build_date" : "2021-07-02T12:06:10.804015202Z",
    "build_snapshot" : false,
    "lucene_version" : "8.8.2",
    "minimum_wire_compatibility_version" : "6.8.0",
    "minimum_index_compatibility_version" : "6.0.0-beta1"
  },
  "tagline" : "You Know, for Search"
}  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   522  100   522    0     0   3195      0 --:--:-- --:--:-- --:--:--  3222

10.1.207.180 | CHANGED | rc=0 >>
{
  "name" : "node-180",
  "cluster_name" : "nc-elasticsearch",
  "cluster_uuid" : "_na_",
  "version" : {
    "number" : "7.13.3",
    "build_flavor" : "default",
    "build_type" : "tar",
    "build_hash" : "5d21bea28db1e89ecc1f66311ebdec9dc3aa7d64",
    "build_date" : "2021-07-02T12:06:10.804015202Z",
    "build_snapshot" : false,
    "lucene_version" : "8.8.2",
    "minimum_wire_compatibility_version" : "6.8.0",
    "minimum_index_compatibility_version" : "6.0.0-beta1"
  },
  "tagline" : "You Know, for Search"
}  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   522  100   522    0     0   3139      0 --:--:-- --:--:-- --:--:--  3163

查看集群状态，可以看到三节点集群已经建立

bash-5.0# ansible all -m shell -a 'curl http://0.0.0.0:39200/_cat/nodes?v'
10.1.207.182 | CHANGED | rc=0 >>
ip           heap.percent ram.percent cpu load_1m load_5m load_15m node.role   master name
10.1.207.181            5          73  17    0.77    0.47     0.37 cdfhilmrstw -      node-181
10.1.207.182            5          35  15    0.64    0.29     0.25 cdfhilmrstw -      node-182
10.1.207.180            5          99  30    3.47    1.92     1.76 cdfhilmrstw *      node-180  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   376  100   376    0     0   3079      0 --:--:-- --:--:-- --:--:--  3107

10.1.207.181 | CHANGED | rc=0 >>
ip           heap.percent ram.percent cpu load_1m load_5m load_15m node.role   master name
10.1.207.181            5          73  17    0.77    0.47     0.37 cdfhilmrstw -      node-181
10.1.207.180            5          99  30    3.47    1.92     1.76 cdfhilmrstw *      node-180
10.1.207.182            5          35  15    0.64    0.29     0.25 cdfhilmrstw -      node-182  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   376  100   376    0     0   3719      0 --:--:-- --:--:-- --:--:--  3722

10.1.207.180 | CHANGED | rc=0 >>
ip           heap.percent ram.percent cpu load_1m load_5m load_15m node.role   master name
10.1.207.181            5          73  17    0.77    0.47     0.37 cdfhilmrstw -      node-181
10.1.207.180            5          99  30    3.47    1.92     1.76 cdfhilmrstw *      node-180
10.1.207.182            5          35  15    0.64    0.29     0.25 cdfhilmrstw -      node-182  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   376  100   376    0     0   9153      0 --:--:-- --:--:-- --:--:--  9400

常用运维命令

启动集群

bash-5.0# ansible all -m shell -a '~/elasticsearch.sh start'

停止集群

bash-5.0# ansible all -m shell -a '~/elasticsearch.sh stop'

Q & A

如何彻底删除 Elasticsearch 集群

A: ~/elasticsearch_uninstall.sh 脚本将 kill Elasticsearch 进程，删除程序文件和所有数据文件

bash-5.0# ansible all -m shell -a '~/elasticsearch_uninstall.sh'
10.1.207.182 | CHANGED | rc=0 >>
Kill Elasticsearch process
Delete Elasticsearch data
Delete Elasticsearch programs
Delete elasticsearch_uninstall.sh

10.1.207.180 | CHANGED | rc=0 >>
Kill Elasticsearch process
Delete Elasticsearch data
Delete Elasticsearch programs
Delete elasticsearch_uninstall.sh

10.1.207.181 | CHANGED | rc=0 >>
Kill Elasticsearch process
Delete Elasticsearch data
Delete Elasticsearch programs
Delete elasticsearch_uninstall.sh