>#Kubernetes DaemonSet讲解 [TOC] 需求来源 ------------ ### 背景问题 <p style=" font-weight: 400; line-height: 1.5; color: #212529; -webkit-tap-highlight-color: transparent; box-sizing: border-box; padding: 0px 20px 20px 20px; border: 1px solid #e9ecef; border-left-width: .25rem; border-radius: .25rem; display: block; border-left-color: #5bc0de;"> 我们可以让每个集群内的节点都运行一个相同的Pod吗？ 如果这样做，以下的问题有什么方式来解决： 1. 我们如何保证每个节点都运行一个Pod？ 2. 如果新节点加入集群，如何感知并部署对应的Pod？ 3. 如果有节点退出，如何删除对应的Pod？ 4. 如果Pod状态异常，如何监控并恢复Pod的状态？ </p> ### DaemonSet 守护进程控制器 <p style=" font-weight: 400; line-height: 1.5; color: #212529; -webkit-tap-highlight-color: transparent; box-sizing: border-box; padding: 0px 20px 20px 20px; border: 1px solid #e9ecef; border-left-width: .25rem; border-radius: .25rem; display: block; border-left-color: #5bc0de;"> DaemonSet 能帮助我们做什么事情？ 1. 保证集群内每一个（或一些节点）都运行一组相同的Pod 2. 跟踪集群节点状态，保证新加入的节点自动创建对应的Pod 3. 跟踪集群节点状态，保证移除的节点删除对应的Pod 4. 跟踪Pod状态，保证每个节点Pod处于运行状态 </p> 用例解读 ------------ ### DaemonSet 语法 <p style=" font-weight: 400; line-height: 1.5; color: #212529; -webkit-tap-highlight-color: transparent; box-sizing: border-box; padding: 0px 20px 20px 20px; border: 1px solid #e9ecef; border-left-width: .25rem; border-radius: .25rem; display: block; border-left-color: #5bc0de;"> 适用场景： 1. 集群存储进程：glusterd，ceph 2. 日志收集进程：fluentd，logstash 3. 需要在每个节点运行的监控收集器 </p> DaemonSet-Demo.yaml ```yaml apiVersion: apps/v1 kind: DaemonSet metadata: name: fluentd-elasticsearch labels: k8s-app: fluentd-logging spec: selector: matchLabels: name: fluentd-elasticsearch template: metadata: labels: name: fluentd-elasticsearch spec: tolerations: - key: node-role.kubernetes.io/master effect: NoSchedule containers: - name: fluentd-elasticsearch image: quay.io/fluentd_elasticsearch/fluentd:v2.5.2 resources: limits: memory: 200Mi requests: cpu: 100m memory: 200Mi volumeMounts: - name: varlog mountPath: /var/log - name: varlibdockercontainers mountPath: /var/lib/docker/containers readOnly: true terminationGracePeriodSeconds: 30 volumes: - name: varlog hostPath: path: /var/log - name: varlibdockercontainers hostPath: path: /var/lib/docker/containers ``` ### 查看DaemonSet状态 ```bash # kubectl get ds NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE fluentd-elasticsearch 1 1 1 1 1 <none> 13s # kubectl get pods NAME READY STATUS RESTARTS AGE fluentd-elasticsearch-b2t7l 1/1 Running 0 19s ``` <p style=" font-weight: 400; line-height: 1.5; color: #212529; -webkit-tap-highlight-color: transparent; box-sizing: border-box; padding: 0px 20px 20px 20px; border: 1px solid #e9ecef; border-left-width: .25rem; border-radius: .25rem; display: block; border-left-color: #5bc0de;"> 状态描述：（本集群公有一个节点[所以大家可以多弄几个节点]） - **DESIRED: 需要的Pod个数** - **UP-TO-DATE: 最新创建的个数** - **CURRENT: 当前已存在的pod个数** - **AVAILABLE: 可用pod个数** - **READY: 就绪的个数** - **NODE SELECTOR: 节点选择标签** </p> ### 更新DaemonSet <p style=" font-weight: 400; line-height: 1.5; color: #212529; -webkit-tap-highlight-color: transparent; box-sizing: border-box; padding: 0px 20px 20px 20px; border: 1px solid #e9ecef; border-left-width: .25rem; border-radius: .25rem; display: block; border-left-color: #5bc0de;"> 更新策略： 1. RollingUpdate: DaemonSet默认更新策略，当更新DaemonSet模板后，老的Pod会被先删除，然后再去创建新的Pod，可以配合健康检查做滚动更新。 2. OnDelete： 当DaemonSet模板更新后，只有手动的删除某一个对应的Pod，此节点Pod才会被更新。 </p> ```bash # 更新镜像 # kubectl set image ds/fluentd-elasticsearch fluentd-elasticsearch=fluent/fluentd:v1.4 daemonset.extensions/fluentd-elasticsearch image updated # kubectl rollout status ds/fluentd-elasticsearch Waiting for daemon set "fluentd-elasticsearch" rollout to finish: 0 out of 1 new pods have been updated... Waiting for daemon set "fluentd-elasticsearch" rollout to finish: 0 of 1 updated pods are available... daemon set "fluentd-elasticsearch" successfully rolled out ``` 架构设计 ------------ <p style=" font-weight: 400; line-height: 1.5; color: #212529; -webkit-tap-highlight-color: transparent; box-sizing: border-box; padding: 0px 20px 20px 20px; border: 1px solid #e9ecef; border-left-width: .25rem; border-radius: .25rem; display: block; border-left-color: #5bc0de;"> 1. DaemonSet Controller 负责根据配置创建Pod 2. DaemonSet Controller 跟踪Job状态，更加配置及时重试Pod或者继续创建 3. DaemonSet Controller 会自动添加affinity & label来跟踪对应的pod，并根据配置在每个节点或者适合的部分节点创建Pod </p>