目錄

pod親和性與反親和性

pod親和性

pod反親和性

?pod狀態(tài)與重啟策略

pod狀態(tài)

pod重啟策略

---

本文主要介紹了pod資源與pod相關(guān)的親和性，以及pod的重啟策略

pod親和性與反親和性

pod親和性（podAffinity）有兩種 1.podaffinity，即聯(lián)系比較緊密的pod更傾向于使用同一個區(qū)域 比如tomcat和nginx這樣資源的利用效率更高

2.podunaffinity，即兩套完全相同，或兩套完全不同功能的服務(wù) 為了不互相影響容災(zāi)效果，或者讓服務(wù)之間不會互相影響，更傾向于不適用同一個區(qū)域

那么如何判斷是不是“同一個區(qū)域”就非常重要

#查看幫助
kubectl explain pods.spec.affinity.podAffinity
preferredDuringSchedulingIgnoredDuringExecution #軟親和性，盡可能在一起
requiredDuringSchedulingIgnoredDuringExecution  #硬親和性，一定要在一起

pod親和性

#硬親和性
kubectl explain pods.spec.affinity.podAffinity.requiredDuringSchedulingIgnoredDuringExecutionlabelSelector        <Object>     #以標(biāo)簽為篩選條件，選擇一組親和的podnamespaceSelector    <Object>     #以命名空間為篩選條件，選擇一組親和的podnamespaces   <[]string>           #確定命名空間的位置topologyKey  <string> -required-  #拓?fù)溥壿嬫I，根據(jù)xx判斷是否是同一位置cat > qinhe-pod1.yaml << EOF
apiVersion: v1 
kind: Pod
metadata:name: qinhe1namespace: defaultlabels:user: ws
spec:containers:- name: qinhe1image: docker.io/library/nginximagePullPolicy: IfNotPresent
EOF
kubectl apply -f qinhe-pod1.yaml    #定義一個初始的pod，后面的pod可以依次為參照echo "
apiVersion: v1
kind: Pod
metadata:name: qinhe2labels:app: app1
spec:containers:- name: qinhe2image: docker.io/library/nginximagePullPolicy: IfNotPresentaffinity:podAffinity:  # 和pod親和性requiredDuringSchedulingIgnoredDuringExecution:- labelSelector:  # 以標(biāo)簽為篩選條件matchExpressions:  # 以表達(dá)式進(jìn)行匹配- {key: user, operator: In, values: ["ws"]}topologyKey: kubernetes.io/hostname
#帶有kubernetes.io/hostname標(biāo)簽相同的被認(rèn)為是同一個區(qū)域，即以主機(jī)名區(qū)分
#標(biāo)簽的node被認(rèn)為是統(tǒng)一位置
" > qinhe-pod2.yaml
kubectl apply -f qinhe-pod2.yamlkubectl get pods -owide #因?yàn)閔ostname node1和node2不同，所以只會調(diào)度到node1
NAME     READY   STATUS    RESTARTS   AGE   IP             NODE           NOMINATED NODE   READINESS GATES
qinhe1   1/1     Running   0          68s   10.10.179.9    ws-k8s-node1   <none>           <none>
qinhe2   1/1     Running   0          21s   10.10.179.10   ws-k8s-node1   <none>           <none>#修改
...topologyKey: beta.kubernetes.io/arch
... #node1和node2這兩個標(biāo)簽都相同
kubectl delete -f qinhe-pod2.yaml
kubectl apply -f qinhe-pod2.yaml
kubectl get pods -owide #再查看時會發(fā)現(xiàn)qinhe2分到了node2
NAME     READY   STATUS    RESTARTS   AGE     IP             NODE           NOMINATED NODE   READINESS GATES
qinhe1   1/1     Running   0          4m55s   10.10.179.9    ws-k8s-node1   <none>           <none>
qinhe2   1/1     Running   0          15s     10.10.234.68   ws-k8s-node2   <none>           <none>#清理環(huán)境
kubectl delete -f qinhe-pod1.yaml
kubectl delete -f qinhe-pod2.yaml

pod反親和性

kubectl explain pods.spec.affinity.podAntiAffinity
preferredDuringSchedulingIgnoredDuringExecution      <[]Object>
requiredDuringSchedulingIgnoredDuringExecution       <[]Object>#硬親和性
#創(chuàng)建qinhe-pod3.yaml
cat > qinhe-pod3.yaml << EOF
apiVersion: v1 
kind: Pod
metadata:name: qinhe3namespace: defaultlabels:user: ws
spec:containers:- name: qinhe3image: docker.io/library/nginximagePullPolicy: IfNotPresent
EOF#創(chuàng)建qinhe-pod4.yaml
echo "
apiVersion: v1
kind: Pod
metadata:name: qinhe4labels:app: app1
spec:containers:- name: qinhe4image: docker.io/library/nginximagePullPolicy: IfNotPresentaffinity:podAntiAffinity:  # 和pod親和性requiredDuringSchedulingIgnoredDuringExecution:- labelSelector:  # 以標(biāo)簽為篩選條件matchExpressions:  # 以表達(dá)式進(jìn)行匹配- {key: user, operator: In, values: ["ws"]}  #表達(dá)式user=wstopologyKey: kubernetes.io/hostname  #以hostname作為區(qū)分是否同個區(qū)域
" > qinhe-pod4.yaml
kubectl apply -f qinhe-pod3.yaml
kubectl apply -f qinhe-pod4.yaml
#分配到了不同的node
kubectl get pods -owide
NAME     READY   STATUS    RESTARTS   AGE   IP             NODE           NOMINATED NODE   READINESS GATES
qinhe3   1/1     Running   0          9s    10.10.179.11   ws-k8s-node1   <none>           <none>
qinhe4   1/1     Running   0          8s    10.10.234.70   ws-k8s-node2   <none>           <none>#修改topologyKey
pod4修改為topologyKey: user
kubectl label nodes ws-k8s-node1 user=xhy
kubectl label nodes ws-k8s-node2 user=xhy
#現(xiàn)在node1和node2都會被pod4識別為同一位置，因?yàn)閚ode的label中user值相同kubectl delete -f qinhe-pod4.yaml
kubectl apply -f qinhe-pod4.yaml
#直接顯示離線
kubectl get pods -owide
NAME     READY   STATUS    RESTARTS   AGE     IP             NODE           NOMINATED NODE   READINESS GATES
qinhe3   1/1     Running   0          9m59s   10.10.179.12   ws-k8s-node1   <none>           <none>
qinhe4   0/1     Pending   0          2s      <none>         <none>         <none>           <none>
#查看日志
Warning  FailedScheduling  74s   default-scheduler  0/4 nodes are available: 2 node(s) didn't match pod anti-affinity rules, 2 node(s) had untolerated taint {node-role.kubernetes.io/control-plane: }. preemption: 0/4 nodes are available: 2 No preemption victims found for incoming pod, 2 Preemption is not helpful for scheduling..#pod反親和性的軟親和性與node親和性的軟親和性同理#清理環(huán)境
kubectl label nodes ws-k8s-node1 user-
kubectl label nodes ws-k8s-node2 user-
kubectl delete -f qinhe-pod3.yaml
kubectl delete -f qinhe-pod4.yaml

?pod狀態(tài)與重啟策略

參考文檔：Pod 的生命周期 | Kubernetes

pod狀態(tài)

1.pending——掛起
（1）正在創(chuàng)建pod，檢查存儲、網(wǎng)絡(luò)、下載鏡像等問題
（2）條件不滿足，比如硬親和性，污點(diǎn)等調(diào)度條件不滿足

2.failed——失敗
至少有一個容器因?yàn)槭《Ｖ?#xff0c;即非0狀態(tài)退出

3.unknown——未知
apiserver連不上node節(jié)點(diǎn)的kubelet，通常是網(wǎng)絡(luò)問題

4.Error——錯誤

5.succeeded——成功
pod所有容器成功終止

6.Unschedulable
pod不能被調(diào)度

7.PodScheduled
正在調(diào)度中

8.Initialized
pod初始化完成

9.ImagePullBackOff
容器拉取失敗

10.evicted
node節(jié)點(diǎn)資源不足

11.CrashLoopBackOff
容器曾經(jīng)啟動，但又異常退出了

pod重啟策略

當(dāng)容器異常時，可以通過設(shè)置RestartPolicy字段，設(shè)置pod重啟策略來對pod進(jìn)行重啟等操作

#查看幫助
kubectl explain pod.spec.restartPolicy
KIND:     Pod
VERSION:  v1
FIELD:    restartPolicy <string>
DESCRIPTION:Restart policy for all containers within the pod. One of Always, OnFailure,Never. Default to Always. More info:<https://kubernetes.io/docs/concepts/workloads/pods/pod-lifecycle/#restart-policy>Possible enum values:- `"Always"`   #只要異常退出，立即自動重啟- `"Never"`    #不會重啟容器- `"OnFailure"`#容器錯誤退出，即退出碼不為0時，則自動重啟#測試Always策略，創(chuàng)建always.yaml
cat > always.yaml << EOF
apiVersion: v1
kind: Pod
metadata:name: always-podnamespace: default
spec:restartPolicy: Alwayscontainers:- name: test-podimage: docker.io/library/tomcatimagePullPolicy: IfNotPresent
EOF
kubectl apply -f always.yaml
kubectl get po #查看狀態(tài)
NAME         READY   STATUS    RESTARTS   AGE
always-pod   1/1     Running   0          22s
#進(jìn)入容器去關(guān)閉容器
kubectl exec -it always-pod -- /bin/bash
shutdown.sh
#查看當(dāng)前狀態(tài)，可以看到always-pod重啟計(jì)數(shù)器為1
kubectl get po
NAME         READY   STATUS    RESTARTS     AGE
always-pod   1/1     Running   1 (5s ago)   70s#測試never策略，創(chuàng)建never.yaml
cat > never.yaml << EOF
apiVersion: v1
kind: Pod
metadata:name: never-podnamespace: default
spec:restartPolicy: Nevercontainers:- name: test-podimage: docker.io/library/tomcatimagePullPolicy: IfNotPresent
EOF
kubectl apply -f never.yaml
kubectl exec -it never-pod -- /bin/bash
shutdown.sh
#不會重啟，狀態(tài)為completed
kubectl get pods | grep never
never-pod    0/1     Completed   0            73s#測試OnFailure策略，創(chuàng)建onfailure.yaml
cat > onfailure.yaml << EOF
apiVersion: v1
kind: Pod
metadata:name: onfailure-podnamespace: default
spec:restartPolicy: OnFailurecontainers:- name: test-podimage: docker.io/library/tomcatimagePullPolicy: IfNotPresent
EOF
kubectl apply -f onfailure.yaml
#進(jìn)去后進(jìn)行異常退出
kubectl exec -it onfailure-pod -- /bin/bash
kill 1
#查看pods狀態(tài)，已經(jīng)重啟
kubectl get po  | grep onfailure
onfailure-pod   1/1     Running     1 (43s ago)   2m11s
#進(jìn)入后進(jìn)行正常退出
kubectl exec -it onfailure-pod -- /bin/bash
shutdown.sh
#查看pods狀態(tài)，沒有重啟，進(jìn)入completed狀態(tài)
kubectl get po  | grep onfailure
onfailure-pod   0/1     Completed   1             3m58s#清理環(huán)境
kubectl delete -f always.yaml
kubectl delete -f never.yaml
kubectl delete -f onfailure.yaml