调整addetcd/addnode/addmaster等脚本2

docs/guide/chrony.md

@@ -35,7 +35,7 @@ server {{ groups.deploy[0] }} iburst
 - 修改 ansible hosts 文件，在 `deploy` 节点配置 `NTP_ENABLED=yes` (默认: no)
 - [可选] 修改 roles/chrony/var/main.yml 中的变量定义，关于文件 roles/chrony/var/main.yml 的由来请看[这里](../setup/config_guide.md)
 
-对于新集群或者新节点，`chrony` 的安装配置已经集成到 `90.setup.yml` `01.prepare.yml` `20.addnode.yml` `21.addmaster.yml` 等脚本中；对于已运行中的集群请执行如下命令进行安装：
+对于新集群或者新节点，`chrony` 的安装配置已经集成到脚本中；对于已运行中的集群请执行如下命令进行安装：
 
 `ansible-playbook /etc/ansible/roles/chrony/chrony.yml `
 

roles/kube-router/tasks/main.yml

@@ -6,7 +6,7 @@
   - /opt/kube/kube-system
 
 - name: 准备配置 kube-router DaemonSet (without IPVS)
-  template: src=kuberouter.yaml.j2 dest=/opt/kube/kube-system/kuberouter.yaml
+  template: src=kuberouter.yaml.j2 dest=/opt/kube/kube-system/kube-router.yaml
     
 - name: 下载cni plugins
   copy: src={{ base_dir }}/bin/{{ item }} dest={{ bin_dir }}/{{ item }} mode=0755
@@ -58,7 +58,7 @@
 
 # 只需单节点执行一次
 - name: 运行 kube-router DaemonSet
-  shell: "{{ bin_dir }}/kubectl apply -f /opt/kube/kube-system/kuberouter.yaml"
+  shell: "{{ bin_dir }}/kubectl apply -f /opt/kube/kube-system/kube-router.yaml"
   run_once: true
 
 # 删除原有cni配置

tools/11.deletcd.yml

@@ -1,5 +1,5 @@
 # remove a etcd member 
-- hosts: deploy
+- hosts: localhost 
   vars_prompt:
   - name: "ETCD_TO_DEL"
     prompt: "which etcd node is about to be deleted?(e.g 192.168.1.1)"
@@ -30,7 +30,6 @@
         dest: "{{ base_dir }}/hosts"
         state: absent
         regexp: '{{ ETCD_NAME.stdout }}'
-      connection: local
       when: "ETCD_NAME.stdout != ''"
   
     - name: delete a etcd member
@@ -44,8 +43,7 @@
       when: "ETCD_ID.stdout != ''"
    
     - name: reconfig and restart the etcd cluster
-      shell: "ansible-playbook /etc/ansible/02.etcd.yml > /tmp/ansible-playbook.log 2>&1"
-      connection: local
+      shell: "ansible-playbook {{ base_dir }}/02.etcd.yml > /tmp/ansible-playbook.log 2>&1"
       when: "ETCD_ID.stdout != ''"
     run_once: true
     # 满足条件才进行删除

tools/change_ip_aio.yml

@@ -7,7 +7,7 @@
 # 3. 检查下修改是否成功，并且能够成功执行 ansible all -m ping
 # 4. 运行本脚本 ansible-playbook /etc/ansible/tools/change_ip_aio.yml
 
-- hosts: deploy    # hosts 角色无所谓，反正allinone所有角色都是同个ip
+- hosts: kube-master    # hosts 角色无所谓，反正allinone所有角色都是同个ip
   tasks:
   - name: 删除一些证书和配置，后面会以新IP重新生成
     file: name={{ item }} state=absent
@@ -16,25 +16,25 @@
     - "/etc/kubernetes/ssl/kubernetes.pem"   # 删除旧master证书
     - "/etc/kubernetes/kubelet.kubeconfig"   # 删除旧kubelet配置文件
 
-- hosts: deploy
+- hosts: kube-master
   roles:
   - deploy
   - etcd
   - kube-master
   - kube-node
 
-- hosts: deploy
+- hosts: kube-master
   tasks:
   - name: 删除老IP地址的node
     shell: "{{ bin_dir }}/kubectl get node |grep -v '{{ inventory_hostname }}'|awk '{print $1}' |xargs {{ bin_dir }}/kubectl delete node"
     ignore_errors: true
 
   - name: 删除原network插件部署
-    shell: "{{ bin_dir }}/kubectl delete -f /opt/kube/kube-system/{{ CLUSTER_NETWORK }}/ || \
-            {{ bin_dir }}/kubectl delete -f /root/local/kube-system/{{ CLUSTER_NETWORK }}/"
+    shell: "{{ bin_dir }}/kubectl delete -f /opt/kube/kube-system/{{ CLUSTER_NETWORK }}.yaml || \
+	    {{ bin_dir }}/kubectl delete -f /opt/kube/kube-ovn/"
     ignore_errors: true
 
-- hosts: deploy
+- hosts: kube-master
   roles:
   - { role: calico, when: "CLUSTER_NETWORK == 'calico'" }
   - { role: cilium, when: "CLUSTER_NETWORK == 'cilium'" }

tools/change_k8s_network.yml

@@ -6,18 +6,16 @@
   - name: 获取所有已经创建的POD信息
     command: "{{ bin_dir }}/kubectl get daemonset -n kube-system"
     register: pod_info
-    delegate_to: "{{ groups.deploy[0] }}"
     run_once: true
 
   - name: 删除原network插件部署
-    shell: "{{ bin_dir }}/kubectl delete -f /opt/kube/kube-system/{{ item }}/"
+    shell: "{{ bin_dir }}/kubectl delete -f /opt/kube/kube-system/{{ item }}.yaml"
     with_items:
     - calico
     - cilium
     - flannel
     - kube-router
     when: 'item in pod_info.stdout'
-    delegate_to: "{{ groups.deploy[0] }}"
     run_once: true
     ignore_errors: true
 
@@ -79,21 +77,6 @@
     - kube-proxy
     ignore_errors: true
 
-- hosts:
-  - lb
-  tasks:
-  - name: 重启lb的keepalived服务
-    service: name=keepalived state=restarted
- 
-  - name: 轮询等待apiserver服务恢复
-    command: "{{ bin_dir }}/kubectl get node"
-    register: result
-    until:    result.rc == 0
-    retries:  5
-    delay: 6
-    delegate_to: "{{ groups.deploy[0] }}"
-    run_once: true
-  
 - hosts:
   - kube-master
   - kube-node
@@ -104,7 +87,7 @@
   - { role: flannel, when: "CLUSTER_NETWORK == 'flannel'" }
   - { role: kube-router, when: "CLUSTER_NETWORK == 'kube-router'" }
 
-- hosts: deploy
+- hosts: kube-node
   tasks:
   # 删除所有运行pod，由controller自动重建
   - name: 重启所有pod
@@ -112,4 +95,3 @@
                 do {{ bin_dir }}/kubectl delete pod --all -n $NS; done;"
     ignore_errors: true
     run_once: true
-

tools/easzctl

@@ -78,7 +78,7 @@ function add-node() {
     sed -i "/\[kube-node/a $1 NEW_NODE=yes" $BASEPATH/hosts
 
     # check if playbook runs successfully
-    ansible-playbook $BASEPATH/tools/20.addnode.yml -e NODE_TO_ADD=$1 || { sed -i "/$1 NEW_NODE=yes/d" $BASEPATH/hosts; return 2; }
+    ansible-playbook $BASEPATH/tools/02.addnode.yml -e NODE_TO_ADD=$1 || { sed -i "/$1 NEW_NODE=yes/d" $BASEPATH/hosts; return 2; }
 
     # save current cluster context if needed
     [ -f "$BASEPATH/.cluster/current_cluster" ] && save_context
@@ -96,7 +96,7 @@ function add-master() {
     sed -i "/\[kube-master/a $1 NEW_MASTER=yes" $BASEPATH/hosts
 
     # check if playbook runs successfully
-    ansible-playbook $BASEPATH/tools/21.addmaster.yml -e NODE_TO_ADD=$1 || { sed -i "/$1 NEW_MASTER=yes/d" $BASEPATH/hosts; return 2; }
+    ansible-playbook $BASEPATH/tools/03.addmaster.yml -e NODE_TO_ADD=$1 || { sed -i "/$1 NEW_MASTER=yes/d" $BASEPATH/hosts; return 2; }
 
     # reconfigure and restart the haproxy service on 'kube-node' nodes
     ansible-playbook $BASEPATH/05.kube-node.yml -t restart_lb || { echo "[ERROR] Failed to restart the haproxy service on 'kube-node' nodes!"; return 2; }
@@ -122,7 +122,7 @@ function add-etcd() {
     sed -i "/\[etcd/a $1 NODE_NAME=$NAME" $BASEPATH/hosts
 
     # check if playbook runs successfully
-    ansible-playbook $BASEPATH/tools/19.addetcd.yml -e NODE_TO_ADD=$1 || { sed -i "/$1 NODE_NAME=$NAME/d" $BASEPATH/hosts; return 2; }
+    ansible-playbook $BASEPATH/tools/01.addetcd.yml -e NODE_TO_ADD=$1 || { sed -i "/$1 NODE_NAME=$NAME/d" $BASEPATH/hosts; return 2; }
 
     # restart apiservers to use the new etcd cluster
     ansible-playbook $BASEPATH/04.kube-master.yml -t restart_master || { echo "[ERROR] Unexpected failures in master nodes!"; return 2; }
@@ -137,7 +137,7 @@ function del-etcd() {
     [[ $1 =~ ^(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})(\.(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})){3}$ ]] || { echo "[ERROR] Invalid ip address!"; return 2; }
 
     # 
-    ansible-playbook $BASEPATH/tools/remove_etcd_node.yml -e ETCD_TO_DEL=$1
+    ansible-playbook $BASEPATH/tools/11.deletcd.yml -e ETCD_TO_DEL=$1 || { echo "[ERROR] Failed to delete etcd node: $1!"; return 2; }
 
     # restart apiservers to use the new etcd cluster
     ansible-playbook $BASEPATH/04.kube-master.yml -t restart_master || { echo "[ERROR] Unexpected failures in master nodes!"; return 2; }