---
- name: Verify /storage is a separate mount (optional)
  hosts: k3s_nodes
  become: true
  tasks:
    - name: Check / and /storage mount sources
      when: k3s_verify_storage_mount | default(false) | bool
      block:
        - name: Get mount source for /
          ansible.builtin.command: findmnt -n -o SOURCE /
          register: mnt_root
          changed_when: false

        - name: Get mount source for /storage
          ansible.builtin.command: findmnt -n -o SOURCE /storage
          register: mnt_storage
          changed_when: false
          failed_when: false

        - name: Assert /storage is mounted on a different device than /
          ansible.builtin.assert:
            that:
              - mnt_storage.rc == 0
              - (mnt_root.stdout | trim | length) > 0
              - (mnt_storage.stdout | trim | length) > 0
              - (mnt_root.stdout | trim) != (mnt_storage.stdout | trim)
            fail_msg: >-
              /storage must be a mount point on a block device different from /.
              See docs/00-04-部署环境说明.md and docs/01-06-节点初始化-ansible-实践.md

- name: Init base system
  hosts: k3s_nodes
  become: true
  tasks:
    # 检查当前节点上 firewalld 的运行状态，供后续条件判断使用
    - name: Check if firewalld is running
      ansible.builtin.command: firewall-cmd --state
      register: firewalld_state
      changed_when: false
      failed_when: false

    # 根据全局 timezone 变量设置系统时区（可选）
    - name: Set timezone
      ansible.builtin.command: timedatectl set-timezone {{ timezone }}
      when: timezone is defined and timezone != ""

    # 安装 k3s 所需的基础工具包（curl、git 等）
    - name: Install basic packages
      ansible.builtin.package:
        name:
          - curl
          - git
        state: present

    # 确保 /etc/hosts 中包含所有 k3s 节点的主机名解析（可选）
    - name: Ensure /etc/hosts has entries for all k3s nodes
      ansible.builtin.lineinfile:
        path: /etc/hosts
        regexp: '^\S+\s+{{ item }}\s*$'
        line: "{{ hostvars[item]['ansible_host'] }} {{ item }}"
        state: present
      loop: "{{ groups['k3s_nodes'] }}"
      when:
        - k3s_manage_hosts | default(true) | bool
        - hostvars[item]['ansible_host'] is defined

    # k3s 所需端口：8472/udp（flannel VXLAN）全部节点；6443/tcp（API）仅 server
    # 必须在安装 k3s 前开放，否则 worker 无法连接、flannel 无法建立 overlay
    # 在所有 k3s 节点上开放 flannel VXLAN 所需的 8472/udp 端口
    - name: Open flannel VXLAN port (8472/udp) on all k3s nodes
      ansible.builtin.command: firewall-cmd --permanent --add-port=8472/udp
      when:
        - k3s_manage_firewalld | default(true) | bool
        - firewalld_state.stdout | default('') == 'running'

    # 在 server 节点上开放 k3s API 端口 6443/tcp
    - name: Open k3s API port (6443/tcp) on server
      ansible.builtin.command: firewall-cmd --permanent --add-port=6443/tcp
      when:
        - k3s_manage_firewalld | default(true) | bool
        - inventory_hostname in groups['k3s_server']
        - firewalld_state.stdout | default('') == 'running'

    # 在完成端口放行后重新加载 firewalld 规则
    - name: Reload firewalld after opening k3s ports
      ansible.builtin.command: firewall-cmd --reload
      when:
        - k3s_manage_firewalld | default(true) | bool
        - firewalld_state.stdout | default('') == 'running'

- name: Install k3s server
  hosts: k3s_server
  become: true
  tasks:
    # 在 server 节点上下载安装并启动 k3s server 进程
    - name: Download and install k3s server
      ansible.builtin.shell: |
        curl -sfL https://get.k3s.io | INSTALL_K3S_EXEC="server --data-dir={{ k3s_data_dir }}" sh -
      args:
        creates: "{{ k3s_data_dir }}/server"

- name: Install k3s agent (workers)
  hosts: k3s_worker
  become: true
  serial: 1   # 逐台安装，减轻并行下载对网络的压力
  tasks:
    # 从首个 server 节点读取集群 token（仅执行一次）
    - name: Read k3s token from first server
      ansible.builtin.slurp:
        src: "{{ k3s_data_dir }}/server/token"
      delegate_to: "{{ groups['k3s_server'][0] }}"
      run_once: true
      register: k3s_token_from_server

    # 在各 worker 节点上保存解码后的 token 供后续安装使用
    - name: Set fact for k3s token on workers
      ansible.builtin.set_fact:
        k3s_token: "{{ k3s_token_from_server.content | b64decode | trim }}"

    # 在每个 worker 节点上下载安装并启动 k3s agent 进程
    - name: Install k3s agent
      ansible.builtin.shell: |
        curl -sfL https://get.k3s.io | K3S_URL=https://{{ k3s_server_ip }}:6443 K3S_TOKEN={{ k3s_token }} INSTALL_K3S_EXEC="agent --data-dir={{ k3s_data_dir }}" sh -
      args:
        creates: "{{ k3s_data_dir }}/agent"
      async: 600
      poll: 15

- name: Configure firewalld baseline for k3s (flannel.1 / cni0 -> trusted)
  hosts: k3s_nodes
  become: true
  tasks:
    # 为 k3s 配置 firewalld 基线：将 flannel.1 / cni0 加入 trusted 区域
    - block:
        # 检查节点上 firewalld 是否可用
        - name: Check if firewalld is available
          ansible.builtin.command: firewall-cmd --state
          register: firewalld_check
          changed_when: false
          failed_when: false

        # 等待 CNI 接口 flannel.1 和 cni0 出现（k3s 启动并创建完成）
        - name: Wait for CNI interfaces (flannel.1, cni0) to appear
          ansible.builtin.shell: |
            for i in $(seq 1 120); do
              ip link show flannel.1 >/dev/null 2>&1 && ip link show cni0 >/dev/null 2>&1 && exit 0
              sleep 1
            done
            exit 1
          when: firewalld_check.stdout == 'running'

        # 将 flannel.1 / cni0 接口加入 firewalld trusted 区域（运行时和永久）
        - name: Add flannel.1 and cni0 to firewalld trusted zone (runtime + permanent)
          ansible.builtin.shell: |
            firewall-cmd --zone=trusted --add-interface={{ item }}
            firewall-cmd --permanent --zone=trusted --add-interface={{ item }}
          loop:
            - flannel.1
            - cni0
          when: firewalld_check.stdout == 'running'

        # 更新 firewalld 配置使新接口规则立即生效
        - name: Reload firewalld
          ansible.builtin.command: firewall-cmd --reload
          when: firewalld_check.stdout == 'running'
      when: k3s_manage_firewalld | default(true) | bool

- name: Configure CoreDNS (IPv4 upstream for ACME)
  hosts: k3s_server
  become: true
  run_once: true
  vars:
    k3s_kubeconfig: /etc/rancher/k3s/k3s.yaml
  tasks:
    - name: Wait for CoreDNS deployment to be ready
      ansible.builtin.shell: |
        KUBECONFIG={{ k3s_kubeconfig }} kubectl rollout status deployment/coredns -n kube-system --timeout=120s
      when: k3s_manage_coredns | default(true) | bool

    - name: Extract CoreDNS Corefile from ConfigMap
      ansible.builtin.shell: |
        KUBECONFIG={{ k3s_kubeconfig }} kubectl get configmap coredns -n kube-system -o jsonpath='{.data.Corefile}' > /tmp/coredns-corefile.txt
      when: k3s_manage_coredns | default(true) | bool

    - name: Patch Corefile forward to IPv4 (avoid IPv6 upstream in Pod network)
      ansible.builtin.replace:
        path: /tmp/coredns-corefile.txt
        regexp: 'forward \. /etc/resolv\.conf'
        replace: 'forward . {{ coredns_forward_servers }}'
      register: coredns_patched
      when: k3s_manage_coredns | default(true) | bool

    - name: Apply patched CoreDNS ConfigMap and restart
      ansible.builtin.shell: |
        KUBECONFIG={{ k3s_kubeconfig }} kubectl create configmap coredns --from-file=Corefile=/tmp/coredns-corefile.txt -n kube-system --dry-run=client -o yaml | KUBECONFIG={{ k3s_kubeconfig }} kubectl apply -f -
        KUBECONFIG={{ k3s_kubeconfig }} kubectl rollout restart deployment/coredns -n kube-system
        KUBECONFIG={{ k3s_kubeconfig }} kubectl rollout status deployment/coredns -n kube-system --timeout=60s
      when:
        - k3s_manage_coredns | default(true) | bool
        - coredns_patched is changed

    - name: Remove temp Corefile
      ansible.builtin.file:
        path: /tmp/coredns-corefile.txt
        state: absent
      when: k3s_manage_coredns | default(true) | bool

- name: 安装后验证 - traefik / nodes / curl
  hosts: k3s_server
  become: true
  run_once: true
  vars:
    k3s_kubeconfig: /etc/rancher/k3s/k3s.yaml
  tasks:
    # 安装后为控制节点打 control-plane 标签（02-05 矩阵 M1 需此标签才能调度），节点名与 inventory 短主机名一致（ylc61～ylc64）
    - name: Label control-plane nodes (k3s 不默认打标，M1 需此标签)
      ansible.builtin.shell: |
        KUBECONFIG={{ k3s_kubeconfig }} kubectl label node {{ item }} node-role.kubernetes.io/control-plane= --overwrite
      loop: "{{ groups['k3s_server'] | default([]) }}"

    # 可选：为工作节点打 worker 标签（02-05 矩阵 M3 需要）
    - name: 可选 - 为工作节点打 worker 标签（02-05 矩阵 M3 需要）
      ansible.builtin.shell: |
        KUBECONFIG={{ k3s_kubeconfig }} kubectl label node {{ item }} node-role.kubernetes.io/worker= --overwrite
      loop: "{{ groups['k3s_worker'] | default([]) }}"
      when: k3s_manage_role_labels | default(true) | bool

    # 查看 kube-system 命名空间中与 Traefik / svclb 相关的 Pod 列表
    - name: kubectl get pods -n kube-system（traefik / svclb）
      ansible.builtin.shell: KUBECONFIG={{ k3s_kubeconfig }} kubectl get pods -n kube-system -o wide | grep -E 'NAME|traefik|svclb'
      register: verify_traefik
      changed_when: false

    # 打印上一步查询到的 Traefik 相关 Pod 信息
    - name: ">>> Traefik 相关 Pods"
      ansible.builtin.debug:
        msg: "{{ item }}"
      loop: "{{ verify_traefik.stdout_lines }}"

    # 查询当前集群中的节点列表
    - name: kubectl get nodes
      ansible.builtin.shell: KUBECONFIG={{ k3s_kubeconfig }} kubectl get nodes
      register: verify_nodes
      changed_when: false

    # 打印节点列表结果，方便确认节点状态与角色
    - name: ">>> kubectl get nodes"
      ansible.builtin.debug:
        msg: "{{ item }}"
      loop: "{{ verify_nodes.stdout_lines }}"

    # 通过 curl 测试每个节点 80 与 443 入口连通性
    - name: curl 测试各节点 80/443 可达性
      ansible.builtin.shell: |
        for ip in {{ groups['k3s_nodes'] | map('extract', hostvars) | map(attribute='ansible_host') | join(' ') }}; do
          c80=$(curl -sk -o /dev/null -w "%{http_code}" --connect-timeout 2 http://$ip 2>/dev/null) || c80="fail"
          c443=$(curl -sk -o /dev/null -w "%{http_code}" --connect-timeout 2 https://$ip 2>/dev/null) || c443="fail"
          echo "$ip: 80=$c80 443=$c443"
        done
      register: verify_curl
      changed_when: false

    - name: ">>> curl 结果"
      ansible.builtin.debug:
        msg: "{{ item }}"
      loop: "{{ verify_curl.stdout_lines }}"