本节将演示如何使用 Ansible 自动化部署 HAProxy 负载均衡器，并结合 Apache Web 服务器实现后端服务的负载均衡。我们将通过配置主机清单、编写 Playbook 以及使用 Jinja2 模板动态生成 HAProxy 配置文件来完成整个部署流程。
配置主机清单
首先，在 Ansible 的主机清单文件 hosts 中定义两个主机组：[webservers] 用于存放后端 Web 服务器，[haproxy] 用于部署负载均衡器。同时，为 webservers 组设置变量 http_port。
ini
[kyy@server1 ansible]$ vim hosts
[webservers]
server2
host1
[webservers:vars]
http_port=80
[haproxy]
server1编写 Keepalived 配置
（此部分为可选的高可用配置示例，用于停止 Keepalived 服务。）
yaml
[kyy@server1 ansible]$ vim keepalived.yml
- name: Start service keepalived
ansible.builtin.service:
name: keepalived
state: stopped
enabled: no
tags: t1编写主部署 Playbook
创建主 Playbook 文件 haproxy.yml，其核心逻辑是：在 haproxy 组的主机上部署 HAProxy，在 webservers 组的主机上部署 Apache。
yaml
[kyy@server1 ansible]$ vim haproxy.yml
---
- hosts: haproxy,webservers
tasks:
- name: deploy haproxy
block:
- name: install haproxy
ansible.builtin.yum:
name: haproxy
state: present
- name: configure the haproxy
ansible.builtin.template:
src: haproxy.cfg.j2
dest: /etc/haproxy/haproxy.cfg
notify: restart service haproxy
- name: enable service haproxy
ansible.builtin.service:
name: haproxy
state: started
enabled: yes
when: ansible_hostname == 'server1'
name: deploy apache
block:
name: Install the Apache
ansible.builtin.yum:
name: httpd
state: present
name: Ensure the default Apache port is {{ http_port }}
ansible.builtin.template:
src: httpd.conf.j2
dest: /etc/httpd/conf/httpd.conf
notify: restart service httpd
name: Start service httpd, if not started
ansible.builtin.service:
name: httpd
state: started
enabled: yes
name: create index.html
ansible.builtin.copy:
content: "{{ ansible_hostname }}\n"
dest: /var/www/html/index.html
when: ansible_hostname in groups['webservers']
handlers:
name: restart service haproxy
ansible.builtin.service:
name: haproxy
state: restarted
name: restart service httpd
ansible.builtin.service:
name: httpd
state: restarted</code></pre>
6.4 准备 HAProxy 配置模板
将现有的 HAProxy 配置文件复制为 Jinja2 模板 haproxy.cfg.j2，并修改其内容，使其能动态引用 webservers 组中所有主机的信息。
[kyy@server1 ansible]$ ansible-playbook haproxy.yml --syntax-check
playbook: haproxy.yml
[kyy@server1 ansible]$ cp /etc/haproxy/haproxy.cfg .
[kyy@server1 ansible]$ mv haproxy.cfg haproxy.cfg.j2
[kyy@server1 ansible]$ vim haproxy.cfg.j2
frontend main
bind *:80
default_backend app
#---------------------------------------------------------------------
round robin balancing between the various backends
#---------------------------------------------------------------------
backend app
balance roundrobin
{% for host in groups['webservers'] %}
server {{ hostvars[host].ansible_hostname }} {{ hostvars[host].ansible_eth0.ipv4.address }}:{{ hostvars[host].http_port }} check
{% endfor %}
6.5 执行 Playbook 并验证
执行 Playbook，Ansible 将根据主机名自动判断任务执行目标：在 server1（haproxy组）上安装并配置 HAProxy，在 server2 和 host1（webservers组）上安装 Apache。
[kyy@server1 ansible]$ ansible-playbook haproxy.yml
PLAY [haproxy,webservers]
TASK [Gathering Facts]
ok: [host1]
ok: [server2]
ok: [server1]
TASK [install haproxy]
skipping: [server2]
skipping: [host1]
ok: [server1]
TASK [configure the haproxy]
skipping: [server2]
skipping: [host1]
changed: [server1]
TASK [enable service haproxy]
skipping: [server2]
skipping: [host1]
changed: [server1]
TASK [Install the Apache]
skipping: [server1]
ok: [host1]
ok: [server2]
TASK [Ensure the default Apache port is {{ http_port }}]
skipping: [server1]
ok: [host1]
ok: [server2]
TASK [Start service httpd, if not started]
skipping: [server1]
ok: [host1]
ok: [server2]
TASK [create index.html]
skipping: [server1]
ok: [host1]
ok: [server2]
RUNNING HANDLER [restart service haproxy]
changed: [server1]
PLAY RECAP
host1 : ok=5 changed=0 unreachable=0 failed=0 skipped=3 rescued=0 ignored=0
server1 : ok=5 changed=3 unreachable=0 failed=0 skipped=4 rescued=0 ignored=0
server2 : ok=5 changed=0 unreachable=0 failed=0 skipped=3 rescued=0 ignored=0
部署完成后，在浏览器中访问负载均衡器的 IP 地址（例如 192.168.223.161:80），HAProxy 将按照轮询（round-robin）策略将请求分发到后端的 Apache 服务器，从而实现负载均衡。编写haproxy.cfg.j2：模板源文件
编辑 Jinja2 模板，在 backend 段添加循环代码，实现自动读取主机组信息生成后端节点。
HAProxy 前端配置：监听本机所有网卡的 80 端口，所有请求默认转发至后端组
app。Jinja2 循环语句：
groups['webservers']获取清单中 webservers 分组全部主机；循环逐个取出主机存入变量 host，利用
hostvars读取对应主机信息，自动生成多条后端 server 配置；
{% endfor %}结束循环。实现无需手动书写每一台后端节点，动态生成 HAProxy 后端服务器列表。
cpp
[kyy@server1 ansible]$ ansible-playbook haproxy.yml --syntax-check
playbook: haproxy.yml
[kyy@server1 ansible]$ cp /etc/haproxy/haproxy.cfg .
[kyy@server1 ansible]$ mv haproxy.cfg haproxy.cfg.j2
[kyy@server1 ansible]$ vim haproxy.cfg.j2
frontend main
bind *:80
default_backend app
#---------------------------------------------------------------------
# round robin balancing between the various backends
#---------------------------------------------------------------------
backend app
balance roundrobin
{% for host in groups['webservers'] %}
server {{ hostvars[host].ansible_hostname }} {{ hostvars[host].ansible_eth0.ipv4.address }}:{{ hostvars[host].http_port }} check
{% endfor %}
运行playbook，成功
java
[kyy@server1 ansible]$ ansible-playbook haproxy.yml
[kyy@server1 ansible]$ ansible-playbook haproxy.yml
PLAY [haproxy,webservers]
TASK [Gathering Facts]
ok: [host1]
ok: [server2]
ok: [server1]
TASK [install haproxy]
skipping: [server2]
skipping: [host1]
ok: [server1]
TASK [configure the haproxy]
skipping: [server2]
skipping: [host1]
changed: [server1]
TASK [enable service haproxy]
skipping: [server2]
skipping: [host1]
changed: [server1]
TASK [Install the Apache]
skipping: [server1]
ok: [host1]
ok: [server2]
TASK [Ensure the default Apache port is {{ http_port }}]
skipping: [server1]
ok: [host1]
ok: [server2]
TASK [Start service httpd, if not started]
skipping: [server1]
ok: [host1]
ok: [server2]
TASK [create index.html]
skipping: [server1]
ok: [host1]
ok: [server2]
RUNNING HANDLER [restart service haproxy]
changed: [server1]
PLAY RECAP
host1 : ok=5 changed=0 unreachable=0 failed=0 skipped=3 rescued=0 ignored=0
server1 : ok=5 changed=3 unreachable=0 failed=0 skipped=4 rescued=0 ignored=0
server2 : ok=5 changed=0 unreachable=0 failed=0 skipped=3 rescued=0 ignored=0浏览器访问192.168.223.161：80
出现的效果是server2，host1轮换出现，说明haproxy部署成功，实现了负载均衡