mgmt/deployment/ansible/playbooks/setup/setup-disk-monitoring.yml

---
- name: 部署 Telegraf 硬盘监控到 Nomad 集群
  hosts: all
  become: yes
  vars:
    # 连接现有的 InfluxDB 2.x + Grafana 监控栈
    influxdb_url: "{{ influxdb_url | default('http://influxdb1.tailnet-68f9.ts.net:8086') }}"
    influxdb_token: "{{ influxdb_token }}"
    influxdb_org: "{{ influxdb_org | default('nomad') }}"
    influxdb_bucket: "{{ influxdb_bucket | default('nomad_monitoring') }}"
    
    # 远程 Telegraf 配置模式（优先）
    use_remote_config: "{{ use_remote_config | default(true) }}"
    telegraf_config_url: "{{ telegraf_config_url | default('') }}"
    
    # 硬盘监控阈值
    disk_usage_warning: 80  # 80% 使用率警告
    disk_usage_critical: 90 # 90% 使用率严重告警
    
    # 监控间隔（秒）
    collection_interval: 30
    
  tasks:
    - name: 显示正在处理的节点
      debug:
        msg: "🔧 正在为节点 {{ inventory_hostname }} 安装硬盘监控"

    - name: 添加 InfluxData 仓库密钥
      apt_key:
        url: https://repos.influxdata.com/influxdata-archive_compat.key
        state: present
      retries: 3
      delay: 5

    - name: 添加 InfluxData 仓库
      apt_repository:
        repo: "deb https://repos.influxdata.com/ubuntu {{ ansible_distribution_release }} stable"
        state: present
        update_cache: yes
      retries: 3
      delay: 5

    - name: 安装 Telegraf
      apt:
        name: telegraf
        state: present
        update_cache: yes
      retries: 3
      delay: 10

    - name: 创建 Telegraf 配置目录
      file:
        path: /etc/telegraf/telegraf.d
        state: directory
        owner: telegraf
        group: telegraf
        mode: '0755'

    - name: 清理旧的 Telegraf 日志文件（节省硬盘空间）
      file:
        path: "{{ item }}"
        state: absent
      loop:
        - /var/log/telegraf
        - /var/log/telegraf.log
      ignore_errors: yes

    - name: 禁用 Telegraf 日志目录创建
      file:
        path: /var/log/telegraf
        state: absent
      ignore_errors: yes

    - name: 创建 Telegraf 环境变量文件
      template:
        src: telegraf-env.j2
        dest: /etc/default/telegraf
        owner: root
        group: root
        mode: '0600'
        backup: yes
      notify: restart telegraf

    - name: 创建 Telegraf systemd 服务文件（支持远程配置）
      template:
        src: telegraf.service.j2
        dest: /etc/systemd/system/telegraf.service
        owner: root
        group: root
        mode: '0644'
        backup: yes
      notify:
        - reload systemd
        - restart telegraf
      when: telegraf_config_url is defined and telegraf_config_url != ''

    - name: 生成 Telegraf 主配置文件（本地配置模式）
      template:
        src: telegraf.conf.j2
        dest: /etc/telegraf/telegraf.conf
        owner: telegraf
        group: telegraf
        mode: '0644'
        backup: yes
      notify: restart telegraf
      when: telegraf_config_url is not defined or telegraf_config_url == ''

    - name: 生成硬盘监控配置
      template:
        src: disk-monitoring.conf.j2
        dest: /etc/telegraf/telegraf.d/disk-monitoring.conf
        owner: telegraf
        group: telegraf
        mode: '0644'
        backup: yes
      notify: restart telegraf

    - name: 生成系统监控配置
      template:
        src: system-monitoring.conf.j2
        dest: /etc/telegraf/telegraf.d/system-monitoring.conf
        owner: telegraf
        group: telegraf
        mode: '0644'
        backup: yes
      notify: restart telegraf

    - name: 启用并启动 Telegraf 服务
      systemd:
        name: telegraf
        state: started
        enabled: yes
        daemon_reload: yes

    - name: 验证 Telegraf 状态
      systemd:
        name: telegraf
      register: telegraf_status

    - name: 检查 InfluxDB 连接
      uri:
        url: "{{ influxdb_url }}/ping"
        method: GET
        timeout: 5
      register: influxdb_ping
      ignore_errors: yes
      delegate_to: localhost
      run_once: true

    - name: 显示 InfluxDB 连接状态
      debug:
        msg: "{{ '✅ InfluxDB 连接正常' if influxdb_ping.status == 204 else '❌ InfluxDB 连接失败，请检查配置' }}"
      run_once: true

    - name: 显示 Telegraf 状态
      debug:
        msg: "✅ Telegraf 状态: {{ telegraf_status.status.ActiveState }}"

    - name: 检查硬盘使用情况
      shell: |
        df -h | grep -vE '^Filesystem|tmpfs|cdrom|udev' | awk '{print $5 " " $1 " " $6}' | while read output;
        do
          usage=$(echo $output | awk '{print $1}' | sed 's/%//g')
          partition=$(echo $output | awk '{print $2}')
          mount=$(echo $output | awk '{print $3}')
          if [ $usage -ge {{ disk_usage_warning }} ]; then
            echo "⚠️  警告: $mount ($partition) 使用率 $usage%"
          else
            echo "✅ $mount ($partition) 使用率 $usage%"
          fi
        done        
      register: disk_check
      changed_when: false

    - name: 显示硬盘检查结果
      debug:
        msg: "{{ disk_check.stdout_lines }}"

  handlers:
    - name: reload systemd
      systemd:
        daemon_reload: yes

    - name: restart telegraf
      systemd:
        name: telegraf
        state: restarted