最新2021全国职业技能大赛---Ansible部署MariaDB高可用

Ansible部署MariaDB高可用

题目：使用OpenStack私有云平台，创建4台系统为centos7.5的云主机，其中一台作为Ansible的母机并命名为ansible，另外三台云主机命名为node1、node2、node3，通过http:///ansible.tar.gz软件包在ansible节点安装Ansible服务；使用这一台母机，编写Ansible脚本（在/root目录下创建example目录作为Ansible工作目录，部署的入口文件命名为cscc_install.yaml），对其他三台云主机进行安装高可用数据库集群（MariaDB_Galera_cluster，数据库密码设置为123456）

1.安装好Ansible，Epel源即可

环境说明：一共四台主机，一台为Ansible母机，其余三台为Ansible操控的子主机：node1，node2，node3

2.设置Ansible的工作目录

题目要求将/root/example目录设置为Ansible的工作目录，(/etc/ansible是Ansible的默认工作目录)，在设置/root/example的工作目录时，可以直接复制默认工作目录的字段

[root@ansible ~]# mkdir example
[root@ansible ~]# cd example/
[root@ansible example]# vi ansible.cfg 
[defaults]
inventory = /root/example/inventory   ###声明主机组
remote_user = root
[root@ansible example]# vim inventory  ###设置主机组，便于后面“when模块”进行筛选使用
[node1]
node1
[node2]
node2
[node3]
node3

测试个node主机和Ansible主机的连通性,可以在Ansible主机上生成个密钥对，把公钥传入到各个node节点，这样方便测试

[root@ansible example]# ansible all -m ping
[WARNING]: Found both group and host with same name: node1
[WARNING]: Found both group and host with same name: node3
[WARNING]: Found both group and host with same name: node2
The authenticity of host 'node1 (192.168.9.130)' can't be established.
ECDSA key fingerprint is SHA256:0dbDTMDX2KJ/jSvS/IIlOPbm3u+AYjSX0OVDyHxtrwQ.
ECDSA key fingerprint is MD5:25:36:7d:13:6c:ab:6b:94:66:d0:cc:a8:f2:3e:d0:9f.
Are you sure you want to continue connecting (yes/no)? 
node2 | SUCCESS => {
    "ansible_facts": {
        "discovered_interpreter_python": "/usr/bin/python"
    }, 
    "changed": false, 
    "ping": "pong"
}
node3 | SUCCESS => {
    "ansible_facts": {
        "discovered_interpreter_python": "/usr/bin/python"
    }, 
    "changed": false, 
    "ping": "pong"
}

node1 | UNREACHABLE! => {
    "changed": false, 
    "msg": "Failed to connect to the host via ssh: Host key verification failed.", 
    "unreachable": true
}
可以看到连通性完好，如果不设置密钥队，需要加个参数"-k"去测试，并且需要输入密码
[root@ansible example]# ansible all -m ping -k
SSH password: 
输入node节点的密码即可检测

3.设置Role目录，编写Playbook

所有的Role目录层级，类似于：
[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-sHIFhdeK-1640089940471)(C:\Users\郑波\AppData\Roaming\Typora\typora-user-images\1636078210464.png)]

[root@ansible example]# mkdir roles    
[root@ansible example]# cd roles/     
[root@ansible roles]# ansible-galaxy init mariadb-galera-cluster        ######生成一个mariadb-galera-cluster角色，并且Role中的文件夹自动生成
- Role mariadb-galera-cluster was created successfully
[root@ansible roles]# ll
total 0
drwxr-xr-x. 10 root root 154 Nov  4 22:11 mariadb-galera-cluster
[root@ansible roles]# cd mariadb-galera-cluster/
[root@ansible mariadb-galera-cluster]# ll
total 4
drwxr-xr-x. 2 root root   22 Nov  4 22:11 defaults
drwxr-xr-x. 2 root root    6 Nov  4 22:11 files
drwxr-xr-x. 2 root root   22 Nov  4 22:11 handlers
drwxr-xr-x. 2 root root   22 Nov  4 22:11 meta
-rw-r--r--. 1 root root 1328 Nov  4 22:11 README.md
drwxr-xr-x. 2 root root   22 Nov  4 22:11 tasks
drwxr-xr-x. 2 root root    6 Nov  4 22:11 templates
drwxr-xr-x. 2 root root   39 Nov  4 22:11 tests
drwxr-xr-x. 2 root root   22 Nov  4 22:11 vars

整体结构构建完毕，下一步：编写tasks

编写tasks的流程：
1.编写yum.yml文件，拷贝安装MariaDB，Galera所需要的安装包

2.编写mysql.yml文件，拷贝mysql.repo文件，

3.编写install.yml文件，安装MariaDb，Galera

4.编写server.yml文件，拷贝/etc/my.cnf.d/server.cnf文件，使得Galera开启，并且也开启MariaDB，并完成数据库的初始化

进入/root/example/roles/mariadb-galera-cluster/tasks目录

3.1：编写yum.yml文件

这一步需要注意，由于在初始化mysql时需要MySQL-python这个包，所以我们在配置Yum包时，要满足能够下载mariaDB-server，galera，MySQL-python(这个包CentOS7.5镜像自带的就有)

[root@ansible tasks]# vi yum.yml
- name: copy mariadb-repo 
  copy: src=/opt/mariadb-repo dest=/opt/

3.2：编写mysql.yml文件

- name: clean .repo
  shell: 'rm -rf /etc/yum.repos.d/*'
- name: copy mysql.repo
  copy: src=/root/mysql.repo dest=/etc/yum.repos.d/mysql.repo
  查看mysql.repo的内容
[ansible]
name=ansible
baseurl=file:///opt/mariadb-repo
enabled=1
gpgcheck=0
[centos]
name=centos
baseurl=ftp://192.168.100.10/centos
enabled=1
gpgcheck=0

3.3：编写install.yml

[root@ansible tasks]# vi install.yml
- name: install mariadb,galera
  yum: name=mariadb-server,galera,MySQL-python

3.4：编写service.yml文件

这一步的前提是需要准备好/etc/my.cnf.d/server.cnf文件

例如：准备node2节点的文件：

[galera]
# Mandatory settings
wsrep_on=ON
wsrep_provider=/usr/lib64/galera/libgalera_smm.so
wsrep_cluster_address="gcomm://192.168.9.131,192.168.9.132"
binlog_format=row
default_storage_engine=InnoDB
innodb_autoinc_lock_mode=2
wsrep_node_name=node2
wsrep_node_address=192.168.9.131
#
# Allow server to accept connections on all interfaces.
#
bind-address=192.168.9.131
#
# Optional setting
wsrep_slave_threads=1
innodb_flush_log_at_trx_commit=0
innodb_buffer_pool_size=120M
wsrep_sst_method=rsync
wsrep_causal_reads=ON
# this is only for embedded server

在准备node1，和node3上的文件，就是把"wsrep_node_name=",“wsrep_node_address=”,"bind-address="这三个字段的值，替换为node1和node3节点的ip和主机名即可

开始编写service.yml文件：

这里需要注意：node1在配置文件传过来之后，需要关闭MariaDBDB服务，node2，node3只需要重启MariaDB服务即可

- name: start node1 mariadb
  service: name=mariadb state=started enabled=yes
  when: "'node1' in group_names"

- name: set node1 mysql_user
  mysql_user: name=root password=123456 state=present
  when: "'node1' in group_names"

- name: start node2 mariadb
  service: name=mariadb state=started enabled=yes
  when: "'node2' in group_names"

- name: set node2 mysql_user
  mysql_user: name=root password=123456 state=present
  when: "'node2' in group_names"

- name: start node3 mariadb
  service: name=mariadb state=started enabled=yes
  when: "'node3' in group_names"

- name: set node3 mysql_user
  mysql_user: name=root password=123456 state=present
  when: "'node3' in group_names"

- name: copy node1 config
  copy: src=/root/node1-server.cnf dest=/etc/my.cnf.d/server.cnf
  when: "'node1' in group_names"

- name: copy node2 config
  copy: src=/root/node2-server.cnf dest=/etc/my.cnf.d/server.cnf
  when: "'node2' in group_names"

- name: copy node3 config
  copy: src=/root/node3-server.cnf dest=/etc/my.cnf.d/server.cnf
  when: "'node3' in group_names"

- name: stop node1 mariadb
  service: name=mariadb state=stopped
  when: "'node1' in group_names"

- name: start node1 galera
  shell: galera_new_cluster
  when: "'node1' in group_names"

- name: restart node2 mariadb
  service: name=mariadb state=restarted
  when: "'node2' in group_names"

- name: restart node3 mariadb
  service: name=mariadb state=restarted
  when: "'node3' in group_names"
  #########这里面有个mysql_user模块，通过“ansible-doc -s mysql_user”来查看这个模块的不详细信息

3.6：编写main.yml

main.yml文件时生成角色时，自动创建好的文件，我们只需要进去编辑

[root@ansible tasks]# vi main.yml 
---
# tasks file for mariadb-galera-cluster
- include: yum.yml
- include: mysql.yml
- include: install.yml
- include: service.yml
注意：main.yml文件一定要按照顺序编写，main.yml是按照顺序执行的

6.编写入口文件

根据题目要求入口文件命名：cscc_install.yaml

注意：此文件需要在/etc/example/目录下，和roles目录在同一级别

[root@ansible example]# vi cscc_install.yaml 
- hosts: all
  remote_user:all
  roles:
  - mariadb-galera-cluster            ######角色名

在来看看此roles的目录层次结构：

[root@ansible example]# tree
.
├── ansible.cfg
├── cscc_install.yaml
├── inventory
└── roles
    └── mariadb-galera-cluster
        ├── defaults
        │   └── main.yml
        ├── files
        ├── handlers
        │   └── main.yml
        ├── meta
        │   └── main.yml
        ├── README.md
        ├── tasks
        │   ├── install.yml
        │   ├── main.yml
        │   ├── mysql.yml
        │   ├── service.yml
        │   └── yum.yml
        ├── templates
        ├── tests
        │   ├── inventory
        │   └── test.yml
        └── vars
            └── main.yml

检查是否有语法错误：

[root@ansible example]# ansible-playbook cscc_install.yaml --syntax-check
[WARNING]: Found both group and host with same name: node1
[WARNING]: Found both group and host with same name: node3
[WARNING]: Found both group and host with same name: node2

playbook: cscc_install.yaml
好像没有

如果没有语法错误，就直接跑playbook

[root@ansible example]# ansible-playbook cscc_install.yaml 

8.后续的检查，测试

在所有子主机上：
ss -ntpul
看看3306，4567端口是否开启
mysql -uroot -p123456
数据库是否能登入进去
在开启了Galera的节点上，进入mysql，执行"show status like 'wsrep%'";
查看集群信息，是否合理
在node1节点上创建个"tets"数据库，在node2，node3上是否有"test"数据库
并插入字段，看各个节点是否有数据

后续更新

会在上面的基础之上部署Haproxy负载均衡案例…