pod 的高階實現汙點親密性探針的實現

烟雨楼台，行云流水發表於2024-10-17

名稱空間資源限制；對名稱空間使用資源的總限額，不是單個pod

[root@master pod]# cat ../nsrq.yaml 
apiVersion: v1
kind: ResourceQuota
metadata:
  name: mem-cpu-quota
  namespace: test
spec:
  hard:
    limits.cpu: "2"
    limits.memory: "2Gi"
    requests.cpu: "4"
    requests.memory: "4Gi"
[root@master pod]# kubectl get po -n test 
No resources found in test namespace.
[root@master pod]# kubectl apply -f ../nsrq.yaml 
resourcequota/mem-cpu-quota configured
[root@master pod]# kubectl describe -n test resourcequotas 
Name:            mem-cpu-quota
Namespace:       test
Resource         Used  Hard
--------         ----  ----
limits.cpu       0     2
limits.memory    0     2Gi
requests.cpu     0     4
requests.memory  0     4Gi
[root@master pod]# cat po
po2.yaml  po3.yaml  po.yaml   
[root@master pod]# cat po.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pod-test
  namespace: test
  labels:
    app: tomcat
spec:
  containers:
  - name: tomcat-test
    image: tomcat
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 8080   
    resources:
      limits:
        memory: "2Gi"
        cpu: "1"
      requests: 
        memory: "2Gi"
        cpu: "1"
[root@master pod]# kubectl apply -f po.yaml 
pod/pod-test created
[root@master pod]# cat po2.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pod-test-1
  namespace: test
  labels:
    app: tomcat
spec:
  containers:
  - name: tomcat-test
    image: tomcat
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 8080   
    resources:
      limits:
        memory: "100Mi"
        cpu: "500m"
      requests: 
        memory: "100Mi"
        cpu: "500m"
[root@master pod]# kubectl apply -f po2.yaml 
Error from server (Forbidden): error when creating "po2.yaml": pods "pod-test-1" is forbidden: exceeded quota: mem-cpu-quota, requested: limits.memory=100Mi, used: limits.memory=2Gi, limited: limits.memory=2Gi
[root@master pod]# kubectl get po -n test 
NAME       READY   STATUS    RESTARTS   AGE
pod-test   1/1     Running   0          2m53s
[root@master pod]# kubectl edit resourcequotas -n test mem-cpu-quota 

# Please edit the object below. Lines beginning with a '#' will be ignored,
# and an empty file will abort the edit. If an error occurs while saving this file will be
# reopened with the relevant failures.
#
apiVersion: v1
kind: ResourceQuota
metadata:
  annotations:
    kubectl.kubernetes.io/last-applied-configuration: |
      {"apiVersion":"v1","kind":"ResourceQuota","metadata":{"annotations":{},"name":"mem-cpu-quota","namespace":"test"},"spec":{"hard":{"limits.cpu":"2","limits.memory":"2Gi","requests.cpu":"4","requests.memor
y":"4Gi"}}}  creationTimestamp: "2024-10-11T07:12:37Z"
  name: mem-cpu-quota
  namespace: test
  resourceVersion: "113002"
  uid: 9f669803-9b56-46cd-a10d-0291b5c59576
spec:
  hard:
    limits.cpu: "2"
    limits.memory: 3Gi # 修改為3G
    requests.cpu: "4"
    requests.memory: 4Gi
status:
  hard:
    limits.cpu: "2"
    limits.memory: 2Gi
    requests.cpu: "4"
    requests.memory: 4Gi
  used:
    limits.cpu: "1"
    limits.memory: 2Gi
    requests.cpu: "1"
    requests.memory: 2Gi
[root@master pod]# kubectl describe -n test resourcequotas 
Name:            mem-cpu-quota
Namespace:       test
Resource         Used  Hard
--------         ----  ----
limits.cpu       1     2
limits.memory    2Gi   3Gi
requests.cpu     1     4
requests.memory  2Gi   4Gi

[root@master pod]# kubectl apply -f po2.yaml 
pod/pod-test-1 created
[root@master pod]# kubectl get  po -n test 
NAME         READY   STATUS    RESTARTS   AGE
pod-test     1/1     Running   0          1s
pod-test-1   1/1     Running   0          11s

　　po 指定node 節點

[root@master pod]# cat po3.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: demo-pod
  namespace: default
  labels:
    app: busybox-tomcat
    env: pro
spec:
  nodeName: node-1
  containers: 
  - name: tomcat
    ports:
    - containerPort: 8080
    image: tomcat:8.5-jre8-alpine
    imagePullPolicy: IfNotPresent
  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command:
    - "/bin/sh"
    - "-c"
    - "sleep 36000"
[root@master pod]# kubectl get po demo-pod -owide 
NAME       READY   STATUS    RESTARTS   AGE    IP              NODE     NOMINATED NODE   READINESS GATES
demo-pod   2/2     Running   0          124m   172.16.84.139   node-1   <none>           <none>

　　po 的透過標籤選擇器選定node

[root@master pod]# cat po4.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: demo-pod-1
  namespace: default
  labels:
    app: busybox-tomcat
    env: pro
spec:
#  nodeName: node-1
  nodeSelector:  
    app: v1  #選擇擁有此標籤的node
  containers: 
  - name: tomcat
    ports:
    - containerPort: 8080
    image: tomcat:8.5-jre8-alpine
    imagePullPolicy: IfNotPresent
  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command:
    - "/bin/sh"
    - "-c"
    - "sleep 36000"



[root@master pod]# kubectl get po 
NAME                       READY   STATUS    RESTARTS   AGE
demo-pod                   2/2     Running   0          126m
demo-pod-1                 0/2     Pending（待排程狀態）   0          3m55s
hamster-65db5bcc68-fm6f2   1/1     Running   0          13h
hamster-65db5bcc68-j8bss   1/1     Running   0          13h
[root@master pod]# kubectl get node --show-labels 
NAME     STATUS   ROLES           AGE   VERSION   LABELS
node     Ready    control-plane   19h   v1.28.2   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=node,kubernetes.io/os=linux,node-role.kubernetes.io/c
ontrol-plane=,node.kubernetes.io/exclude-from-external-load-balancers=node-1   Ready    <none>          19h   v1.28.2   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=node-1,kubernetes.io/os=linux
node-2   Ready    <none>          19h   v1.28.2   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=node-2,kubernetes.io/os=linux
[root@master pod]# kubectl label nodes node-2 app=v1
node/node-2 labeled
[root@master pod]# kubectl get po 
NAME                       READY   STATUS              RESTARTS   AGE
demo-pod                   2/2     Running             0          128m
demo-pod-1                 0/2     ContainerCreating   0          5m34s
hamster-65db5bcc68-fm6f2   1/1     Running             0          13h
hamster-65db5bcc68-j8bss   1/1     Running             0          13h
[root@master pod]# kubectl get po -owide
NAME                       READY   STATUS              RESTARTS   AGE     IP              NODE     NOMINATED NODE   READINESS GATES
demo-pod                   2/2     Running             0          128m    172.16.84.139   node-1   <none>           <none>
demo-pod-1                 0/2     ContainerCreating   0          5m49s   <none>          node-2   <none>           <none>
hamster-65db5bcc68-fm6f2   1/1     Running             0          13h     172.16.247.10   node-2   <none>           <none>
hamster-65db5bcc68-j8bss   1/1     Running             0          13h     172.16.84.138   node-1   <none>           <none>

　　po 節點硬親和性

[root@master pod]# cat po.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pod-test
#  namespace: test
  labels:
    app: tomcat
spec:
  affinity:
    nodeAffinity: 
      requiredDuringSchedulingIgnoredDuringExecution:  #硬親和性
        nodeSelectorTerms:
        - matchExpressions:
          - key: app  #標籤
            operator: In # 等於
            values:
            - v2 #值 
            - v3
  containers:
  - name: tomcat-test
    image: tomcat
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 8080   
[root@master pod]# kubectl apply -f po.yaml 
pod/pod-test created
[root@master pod]# kubectl get po
NAME                       READY   STATUS    RESTARTS   AGE
demo-pod                   2/2     Running   0          145m
demo-pod-1                 2/2     Running   0          22m
hamster-65db5bcc68-fm6f2   1/1     Running   0          13h
hamster-65db5bcc68-j8bss   1/1     Running   0          13h
pod-test                   0/1     Pending   0          4s
[root@master pod]# kubectl label nodes node-1 app=v2 && kubectl get pod pod-test -owide 
node/node-1 labeled
NAME       READY   STATUS              RESTARTS   AGE   IP       NODE     NOMINATED NODE   READINESS GATES
pod-test   0/1     ContainerCreating   0          50s   <none>   node-1   <none>           <none>

　　po 節點軟親和性

[root@master pod]# cat po6.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pod-test-cx
#  namespace: test
  labels:
    app: tomcat
spec:
  affinity:
    nodeAffinity: 
      preferredDuringSchedulingIgnoredDuringExecution:  #軟親和性
      - preference:
          matchExpressions:
          - key: cx  #標籤
            operator: In # 等於
            values:
            - dev #值 
        weight: 90 # 親和力度
  containers:
  - name: tomcat-test
    image: tomcat
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 8080   
[root@master pod]# kubectl apply -f po6.yaml 
pod/pod-test-cx created
[root@master pod]# kubectl get pod pod-test-cx -owide
NAME          READY   STATUS    RESTARTS   AGE   IP              NODE     NOMINATED NODE   READINESS GATES
pod-test-cx   1/1     Running   0          26s   172.16.247.13   node-2   <none>           <none>
[root@master pod]# kubectl get nodes --show-labels | grep cx=dev

　　打標籤刪除標籤

打標籤
[root@master pod]# kubectl label nodes node-1 app=v2
node/node-1 labeled

刪除標籤
[root@master pod]# kubectl label nodes node-2 app-
node/node-2 unlabeled
[root@master pod]# kubectl label nodes node-1 app-
node/node-1 unlabeled
[root@master pod]# kubectl label nodes node-1 cx-
node/node-1 unlabeled

　　po 與po 的親和性

[root@master pod]# kubectl get po  --show-labels 
NAME       READY   STATUS    RESTARTS   AGE   LABELS
pod-test   1/1     Running   0          94s   app=tomcat,security=S1
[root@master pod]# cat poaffinity.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pod-test-cx
#  namespace: test
  labels:
    app: tomcat
spec:
  affinity:
    podAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:
      requiredDuringSchedulingIgnoredDuringExecution: # 硬親和性
      - labelSelector:
          matchExpressions:
          - key: security
            operator: In
            values:
            - S1
        topologyKey: kubernetes.io/hostname # 根據這個node 標籤判斷是不是一個節點
  containers:
  - name: tomcat-test
    image: tomcat
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 8080   

[root@master pod]# kubectl apply -f poaffinity.yaml 
pod/pod-test-cx created
[root@master pod]# kubectl get po  --show-labels 
NAME          READY   STATUS    RESTARTS   AGE    LABELS
pod-test      1/1     Running   0          109s   app=tomcat,security=S1
pod-test-cx   1/1     Running   0          3s     app=tomcat
[root@master pod]# kubectl get po  -owide
NAME          READY   STATUS    RESTARTS   AGE   IP              NODE     NOMINATED NODE   READINESS GATES
pod-test      1/1     Running   0          2m    172.16.84.147   node-1   <none>           <none>
pod-test-cx   1/1     Running   0          14s   172.16.84.149   node-1   <none>           <none>

　　po反親和性

[root@master pod]# cat nopo.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pod-test-cx
#  namespace: test
  labels:
    app: tomcat
spec:
  affinity:
    podAntiAffinity: # 反親和性
      preferredDuringSchedulingIgnoredDuringExecution:
      requiredDuringSchedulingIgnoredDuringExecution: # 硬反親和性
      - labelSelector:
          matchExpressions:
          - key: security
            operator: In
            values:
            - S1
        topologyKey: kubernetes.io/hostname # 根據這個node 標籤判斷
  containers:
  - name: tomcat-test
    image: tomcat
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 8080   

[root@master pod]# kubectl get po --show-labels 
NAME          READY   STATUS    RESTARTS   AGE     LABELS
pod-test      1/1     Running   0          8m53s   app=tomcat,security=S1
pod-test-cx   1/1     Running   0          18s     app=tomcat
[root@master pod]# kubectl get po -owide 
NAME          READY   STATUS    RESTARTS   AGE   IP              NODE     NOMINATED NODE   READINESS GATES
pod-test      1/1     Running   0          9m    172.16.84.150   node-1   <none>           <none>
pod-test-cx   1/1     Running   0          25s   172.16.247.14   node-2   <none>           <none>

　　汙點與容忍度

KIND:       Node
VERSION:    v1

FIELD: taints <[]Taint>

DESCRIPTION:
    If specified, the node's taints.
    The node this Taint is attached to has the "effect" on any pod that does not
    tolerate the Taint.
    
FIELDS:
  effect	<string> -required-
    Required. The effect of the taint on pods that do not tolerate the taint.
    Valid effects are NoSchedule, PreferNoSchedule and NoExecute.
    
    Possible enum values:
     - `"NoExecute"` Evict any already-running pods that do not tolerate the
    taint. Currently enforced by NodeController.
     - `"NoSchedule"` Do not allow new pods to schedule onto the node unless
    they tolerate the taint, but allow all pods submitted to Kubelet without
    going through the scheduler to start, and allow all already-running pods to
    continue running. Enforced by the scheduler.
     - `"PreferNoSchedule"` Like TaintEffectNoSchedule, but the scheduler tries
    not to schedule new pods onto the node, rather than prohibiting new pods
    from scheduling onto the node entirely. Enforced by the scheduler.

  key	<string> -required-
    Required. The taint key to be applied to a node.

  timeAdded	<string>
    TimeAdded represents the time at which the taint was added. It is only
    written for NoExecute taints.

  value	<string>
    The taint value corresponding to the taint key.

effect 欄位的允許值包括：

NoExecute

這會影響已在節點上執行的 Pod，具體影響如下：

如果 Pod 不能容忍這類汙點，會馬上被驅逐。
如果 Pod 能夠容忍這類汙點，但是在容忍度定義中沒有指定 tolerationSeconds，則 Pod 還會一直在這個節點上執行。
如果 Pod 能夠容忍這類汙點，而且指定了 tolerationSeconds，則 Pod 還能在這個節點上繼續執行這個指定的時間長度。這段時間過去後，節點生命週期控制器從節點驅除這些 Pod。

NoSchedul
除非具有匹配的容忍度規約，否則新的 Pod 不會被排程到帶有汙點的節點上。當前正在節點上執行的 Pod 不會被驅逐。

PreferNoSchedule

PreferNoSchedule 是“偏好”或“軟性”的 NoSchedule。控制平面將嘗試避免將不能容忍汙點的 Pod 排程到的節點上，但不能保證完全避免。

檢視node-1 .節點pod
[root@master ~]# kubectl get node 
NAME     STATUS   ROLES           AGE     VERSION
node     Ready    control-plane   4d18h   v1.28.2
node-1   Ready    <none>          4d18h   v1.28.2
node-2   Ready    <none>          4d18h   v1.28.2
[root@master ~]# kubectl get pod -A -owide | grep node-1
default       pod-test                                    1/1     Running   1 (28m ago)   3d22h   172.16.84.151    node-1   <none>           <none>
kube-system   calico-node-2bjbd                           1/1     Running   3 (28m ago)   4d18h   192.168.10.30    node-1   <none>           <none>
kube-system   kube-proxy-jxl5j                            1/1     Running   3 (28m ago)   4d18h   192.168.10.30    node-1   <none>           <none>
kube-system   vpa-admission-controller-6cfd4f784d-w8c58   1/1     Running   1 (28m ago)   4d17h   172.16.84.153    node-1   <none>           <none>
kube-system   vpa-updater-cc89b6c56-grq55                 1/1     Running   1 (28m ago)   4d17h   172.16.84.152    node-1   <none>           <none>
給node-1 打汙點，並檢視node-1 上pod
[root@master ~]# kubectl taint node node-1 a=b:NoExecute &&  kubectl get pod -A -owide | grep node-1
node/node-1 tainted
default       pod-test                                    1/1     Terminating   1 (30m ago)   3d22h   172.16.84.151    node-1   <none>           <none>
kube-system   calico-node-2bjbd                           1/1     Running       3 (30m ago)   4d18h   192.168.10.30    node-1   <none>           <none>
kube-system   kube-proxy-jxl5j                            1/1     Running       3 (30m ago)   4d19h   192.168.10.30    node-1   <none>           <none>
kube-system   vpa-admission-controller-6cfd4f784d-w8c58   1/1     Terminating   1 (30m ago)   4d17h   172.16.84.153    node-1   <none>           <none>
kube-system   vpa-updater-cc89b6c56-grq55                 1/1     Terminating   1 (30m ago)   4d17h   172.16.84.152    node-1   <none>           <none>

operator 的預設值是 Equal。

一個容忍度和一個汙點相“匹配”是指它們有一樣的鍵名和效果，並且：

如果 operator 是 Exists（此時容忍度不能指定 value），或者
如果 operator 是 Equal，則它們的值應該相等

[root@master ~]# kubectl taint node node-1 a=b:NoSchedule
node/node-1 tainted
[root@master ~]# cat po.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pod-test-1
  labels:
    app: tomcat
spec:
  tolerations:
  - key: a
    operator: Exists  #可以不知道value
    effect: "NoSchedule" #容忍度
  nodeName: node-1
  containers:
  - name: tomcat-test
    image: tomcat
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 8080

　　指定value

[root@master ~]# kubectl apply -f po-2.yaml 
pod/pod-test-2 created
[root@master ~]# kubectl get pod -A -owide | grep node-1
default       pod-test-1                                  1/1     Running   0             4m57s   172.16.84.154    node-1   <none>           <none>
default       pod-test-2                                  1/1     Running   0             4s      172.16.84.155    node-1   <none>           <none>
kube-system   calico-node-2bjbd                           1/1     Running   3 (54m ago)   4d19h   192.168.10.30    node-1   <none>           <none>
kube-system   kube-proxy-jxl5j                            1/1     Running   3 (54m ago)   4d19h   192.168.10.30    node-1   <none>           <none>
[root@master ~]# cat po-2.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pod-test-2
  labels:
    app: tomcat
spec:
  tolerations:
  - key: a
    operator: Equal  #可以省略，預設需要指定value
    value: b
    effect: "NoSchedule"
  nodeName: node-1
  containers:
  - name: tomcat-test
    image: tomcat
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 8080

pod 的探測

1.存活探測： livenessProbe 用於存活探針決定何時重啟容器。檢測pod 容器是否處於執行狀態，當存活性探測失敗。存活探針失敗多次，kubelet 將重啟該容器存活探針不會等待就緒探針成功。如果你想在執行存活探針前等待，你可以定義 initialDelaySeconds，或者使用啟動探針

KIND: Pod
VERSION: v1

FIELD: livenessProbe <Probe>

DESCRIPTION:
Periodic probe of container liveness. Container will be restarted if the
probe fails. Cannot be updated. More info:
https://kubernetes.io/docs/concepts/workloads/pods/pod-lifecycle#container-probes
Probe describes a health check to be performed against a container to
determine whether it is alive or ready to receive traffic.

FIELDS:
exec <ExecAction> # 命令列探測方式
Exec specifies the action to take.

failureThreshold <integer>
Minimum consecutive failures for the probe to be considered failed after
having succeeded. Defaults to 3. Minimum value is 1.

grpc <GRPCAction> #grpc 探測
GRPC specifies an action involving a GRPC port.

httpGet <HTTPGetAction> # http 協議探測
HTTPGet specifies the http request to perform.

initialDelaySeconds <integer> # initialDelaySeconds 欄位告訴 kubelet 在執行第一次探測前應該等待 時間 秒
Number of seconds after the container has started before liveness probes are
initiated. More info:
https://kubernetes.io/docs/concepts/workloads/pods/pod-lifecycle#container-probes

periodSeconds <integer> #periodSeconds 欄位指定了 kubelet 應該每 5 秒執行一次存活探測
How often (in seconds) to perform the probe. Default to 10 seconds. Minimum
value is 1.

successThreshold <integer> #探測成功的最少連續成功次數
Minimum consecutive successes for the probe to be considered successful
after having failed. Defaults to 1. Must be 1 for liveness and startup.
Minimum value is 1.

tcpSocket <TCPSocketAction> # tcp 探測
TCPSocket specifies an action involving a TCP port.

terminationGracePeriodSeconds <integer> # 正常終止的可選持續時間；Pod 中執行的程序會收到終止訊號以及終止時間
Optional duration in seconds the pod needs to terminate gracefully upon
probe failure. The grace period is the duration in seconds after the
processes running in the pod are sent a termination signal and the time when
the processes are forcibly halted with a kill signal. Set this value longer
than the expected cleanup time for your process. If this value is nil, the
pod's terminationGracePeriodSeconds will be used. Otherwise, this value
overrides the value provided by the pod spec. Value must be non-negative
integer. The value zero indicates stop immediately via the kill signal (no
opportunity to shut down). This is a beta field and requires enabling
ProbeTerminationGracePeriod feature gate. Minimum value is 1.
spec.terminationGracePeriodSeconds is used if unset.

timeoutSeconds <integer> #探測超時後的秒數。預設為 1 秒。
Number of seconds after which the probe times out. Defaults to 1 second.
Minimum value is 1. More info:
https://kubernetes.io/docs/concepts/workloads/pods/pod-lifecycle#container-probes

　　http 探測

[root@master pod]# cat livenessProbe-1.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: liveness-1
spec:
  containers:
  - name: liveness
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - name: web
      containerPort: 80
    livenessProbe:
      initialDelaySeconds: 10
      httpGet:
        port: web
        path: /index.html
[root@master pod]# kubectl apply -f livenessProbe-1.yaml 
pod/liveness-1 created
[root@master pod]# kubectl get po
NAME          READY   STATUS              RESTARTS      AGE
liveness-1    0/1     ContainerCreating   0             7s
pod-test-1    1/1     Running             0             159m
pod-test-2    1/1     Running             0             154m
pod-test-cx   1/1     Running             1 (15h ago)   4d1h
[root@master pod]# kubectl get po
NAME          READY   STATUS              RESTARTS      AGE
liveness-1    0/1     ContainerCreating   0             8s
pod-test-1    1/1     Running             0             159m
pod-test-2    1/1     Running             0             154m
pod-test-cx   1/1     Running             1 (15h ago)   4d1h
[root@master pod]# kubectl get po -w
NAME          READY   STATUS              RESTARTS      AGE
liveness-1    0/1     ContainerCreating   0             11s
pod-test-1    1/1     Running             0             159m
pod-test-2    1/1     Running             0             154m
pod-test-cx   1/1     Running             1 (15h ago)   4d1h
liveness-1    1/1     Running             0             29s

　　exec 探測

[root@master pod]# cat livenessProbe-2.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: liveness-2
spec:
  containers:
  - name: liveness
    image: busybox
    imagePullPolicy: IfNotPresent
    command:
    - /bin/sh
    - "-c"
    - "echo 1 > /tmp/1.txt && sleep 120;rm -f /tmp/1.txt;sleep 1200"
    livenessProbe:
      initialDelaySeconds: 10
      exec: 
        command:
        - /bin/sh
        - "-c"
        - "cat /tmp/1.txt"
      periodSeconds: 3
      successThreshold: 1
[root@master pod]# kubectl apply -f livenessProbe-2.yaml && kubectl get po -w
pod/liveness-2 created
NAME          READY   STATUS              RESTARTS      AGE
liveness-1    1/1     Running             0             24m
liveness-2    0/1     ContainerCreating   0             0s
pod-test-1    1/1     Running             0             3h3m
pod-test-2    1/1     Running             0             179m
pod-test-cx   1/1     Running             1 (15h ago)   4d1h
liveness-2    0/1     ContainerCreating   0             1s
liveness-2    1/1     Running             0             2s
liveness-2    1/1     Running             1 (1s ago)    2m41s

　　tcp

[root@master pod]# cat livenessProbe-3.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: liveness-3
spec:
  containers:
  - name: liveness
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - name: web
      containerPort: 80
    livenessProbe:
      initialDelaySeconds: 10
      tcpSocket:
        port: web
      successThreshold: 1
[root@master pod]# kubectl apply -f livenessProbe-3.yaml && kubectl get po -w
pod/liveness-3 created
NAME          READY   STATUS              RESTARTS      AGE
liveness-1    1/1     Running             0             44m
liveness-3    0/1     ContainerCreating   0             0s
pod-test-1    1/1     Running             0             3h23m
pod-test-2    1/1     Running             0             3h18m
pod-test-cx   1/1     Running             1 (16h ago)   4d1h
liveness-3    0/1     ContainerCreating   0             0s
liveness-3    1/1     Running             0             1s

2.就緒探測：就緒探針決定何時容器準備好開始接受流量。這種探針在等待應用執行耗時的初始任務時非常有用，例如建立網路連線、載入檔案和預熱快取。

tcp

[root@master pod]# cat readinessProbe-1.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: periodseconds-1
spec:
  containers:
  - name: liveness
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - name: web
      containerPort: 80
    readinessProbe:
      initialDelaySeconds: 10
      tcpSocket:
        port: web
      successThreshold: 1
      periodSeconds: 3
[root@master pod]# vim readinessProbe-1.yaml 
[root@master pod]# kubectl apply -f readinessProbe-1.yaml && kubectl get po -w
pod/periodseconds-1 created
NAME              READY   STATUS              RESTARTS      AGE
liveness-1        1/1     Running             0             52m
liveness-3        1/1     Running             0             8m54s
periodseconds-1   0/1     ContainerCreating   0             0s
pod-test-1        1/1     Running             0             3h32m
pod-test-2        1/1     Running             0             3h27m
pod-test-cx       1/1     Running             1 (16h ago)   4d1h
periodseconds-1   0/1     ContainerCreating   0             1s
periodseconds-1   0/1     Running             0             1s
periodseconds-1   1/1     Running             0             12s

　　http

[root@master pod]#  cat readinessProbe-2.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: periodseconds-2
spec:
  containers:
  - name: liveness
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - name: web
      containerPort: 80
    readinessProbe:
      initialDelaySeconds: 10
      httpGet:
        path: /index.html
        port: web
      successThreshold: 1
      periodSeconds: 3
[root@master pod]# kubectl apply -f readinessProbe-2.yaml ; kubectl get pod -w 
pod/periodseconds-2 created
NAME              READY   STATUS              RESTARTS      AGE
liveness-1        1/1     Running             0             4h8m
liveness-3        1/1     Running             0             3h24m
periodseconds-1   1/1     Running             0             3h15m
periodseconds-2   0/1     ContainerCreating   0             0s
pod-test-1        1/1     Running             0             6h47m
pod-test-2        1/1     Running             0             6h42m
pod-test-cx       1/1     Running             1 (19h ago)   4d5h
periodseconds-2   0/1     ContainerCreating   0             1s
periodseconds-2   0/1     Running             0             2s
periodseconds-2   1/1     Running             0             12s

　　exec

[root@master pod]# cat readinessProbe-3.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: riodseconds-2
  labels: 
    app: exec
spec:
  containers:
  - name: liveness
    image: nginx
    imagePullPolicy: IfNotPresent
    command: 
    - /bin/sh
    - "-c"
    - "sleep 120;rm -f /usr/share/nginx/html/index.html"
    ports:
    - name: web
      containerPort: 80
    readinessProbe:
      initialDelaySeconds: 10
      exec:
        command:
        - /bin/sh
        - "-c"
        - "ls /usr/share/nginx/html/index.html"
      successThreshold: 1
      periodSeconds: 3
---
apiVersion: v1
kind: Service
metadata:
  name: riodseconds-exec
spec:
  selector:
    app: exec
  ports:
    name: web
    protocol: TCP
    targetPort: web
    port: 80
[root@master pod]# vim readinessProbe-3.yaml 
[root@master pod]# kubectl apply -f readinessProbe-3.yaml ; kubectl get pod && kubectl get svc -w
pod/riodseconds-2 created
service/riodseconds-exec created
NAME              READY   STATUS              RESTARTS      AGE
liveness-1        1/1     Running             0             4h28m
liveness-3        1/1     Running             0             3h43m
periodseconds-1   1/1     Running             0             3h35m
periodseconds-2   1/1     Running             0             19m
pod-test-1        1/1     Running             0             7h7m
pod-test-2        1/1     Running             0             7h2m
pod-test-cx       1/1     Running             1 (19h ago)   4d5h
riodseconds-2     0/1     ContainerCreating   0             0s
NAME               TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)   AGE
kubernetes         ClusterIP   10.96.0.1       <none>        443/TCP   5d2h
riodseconds-exec   ClusterIP   10.108.185.32   <none>        80/TCP    0s
^C[root@master pod]
[root@master pod]# 
[root@master pod]# 
[root@master pod]# 
[root@master pod]# kubectl describe svc riodseconds-exec 
Name:              riodseconds-exec
Namespace:         default
Labels:            <none>
Annotations:       <none>
Selector:          app=exec
Type:              ClusterIP
IP Family Policy:  SingleStack
IP Families:       IPv4
IP:                10.108.185.32
IPs:               10.108.185.32
Port:              web  80/TCP
TargetPort:        web/TCP
Endpoints:         172.16.84.166:80
Session Affinity:  None
Events:            <none>
[root@master pod]# kubectl describe svc riodseconds-exec 
Name:              riodseconds-exec
Namespace:         default
Labels:            <none>
Annotations:       <none>
Selector:          app=exec
Type:              ClusterIP
IP Family Policy:  SingleStack
IP Families:       IPv4
IP:                10.108.185.32
IPs:               10.108.185.32
Port:              web  80/TCP
TargetPort:        web/TCP
Endpoints:         
Session Affinity:  None
Events:            <none>

3.啟動探針檢查容器內的應用是否已啟動。啟動探針可以用於對慢啟動容器進行存活性檢測，避免它們在啟動執行之前就被 kubelet 殺掉。如果配置了這類探針，它會禁用存活檢測和就緒檢測，直到啟動探針成功為止。這類探針僅在啟動時執行，不像存活探針和就緒探針那樣週期性地執行。

　　 exec

[root@master pod]# cat stat-1.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: stat-1
spec:
  containers:
  - name: stat-1
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - name: web
      containerPort: 80
   # readinessProbe:
   #  initialDelaySeconds: 10
   #   httpGet:
   #     path: /index.html
   #     port: web
   #   successThreshold: 1
   #   periodSeconds: 3
    startupProbe:
      #httpGet:
      #  path: /healthz
      #  port: liveness-port
      exec:
        command:
        - /bin/sh
        - "-c"
        - "ls /usr/share/nginx/html/index.html"
      failureThreshold: 3 # 失敗的次數
      periodSeconds: 10 # 探測間隔
[root@master pod]# kubectl apply -f stat-1.yaml ; kubectl get po -w
pod/stat-1 created
NAME              READY   STATUS              RESTARTS      AGE
liveness-1        1/1     Running             0             7h11m
liveness-3        1/1     Running             0             6h27m
periodseconds-1   1/1     Running             0             6h18m
periodseconds-2   1/1     Running             0             3h3m
pod-test-1        1/1     Running             0             9h
pod-test-2        1/1     Running             0             9h
pod-test-cx       1/1     Running             1 (22h ago)   4d8h
stat-1            0/1     ContainerCreating   0             0s
stat-1            0/1     ContainerCreating   0             0s
stat-1            0/1     Running             0             1s
stat-1            0/1     Running             0             10s
stat-1            1/1     Running             0             10s

　　tcp

[root@master pod]# cat stat-2.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: stat-2
spec:
  containers:
  - name: stat-2
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - name: web
      containerPort: 80
   # readinessProbe:
   #  initialDelaySeconds: 10
   #   httpGet:
   #     path: /index.html
   #     port: web
   #   successThreshold: 1
   #   periodSeconds: 3
    startupProbe:
      #httpGet:
      #  path: /healthz
      #  port: liveness-port
      #exec:
      #  command:
      #  - /bin/sh
      #  - "-c"
      #  - "ls /usr/share/nginx/html/index.html"
      tcpSocket:
        port: web
      failureThreshold: 3 # 失敗的次數
      periodSeconds: 10 # 探測間隔
[root@master pod]# kubectl apply -f stat-2.yaml ; kubectl get po -w
pod/stat-2 created
NAME              READY   STATUS              RESTARTS      AGE
liveness-1        1/1     Running             0             19h
liveness-3        1/1     Running             0             18h
periodseconds-1   1/1     Running             0             18h
periodseconds-2   1/1     Running             0             15h
pod-test-1        1/1     Running             0             22h
pod-test-2        1/1     Running             0             22h
pod-test-cx       1/1     Running             1 (35h ago)   4d20h
stat-1            1/1     Running             0             12h
stat-2            0/1     ContainerCreating   0             0s
stat-2            0/1     ContainerCreating   0             1s
stat-2            0/1     Running             0             1s
stat-2            0/1     Running             0             10s
stat-2            1/1     Running             0             11s

　　http

[root@master pod]# cat stat-3.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: stat-3
spec:
  containers:
  - name: stat-3
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - name: web
      containerPort: 80
   # readinessProbe:
   #  initialDelaySeconds: 10
   #   httpGet:
   #     path: /index.html
   #     port: web
   #   successThreshold: 1
   #   periodSeconds: 3
    startupProbe:
      httpGet:
        path: /index.html
        port: web
      #exec:
      #  command:
      #  - /bin/sh
      #  - "-c"
      #  - "ls /usr/share/nginx/html/index.html"
      #tcpSocket:
      #  port: web
      failureThreshold: 3 # 失敗的次數
      periodSeconds: 10 # 探測間隔
[root@master pod]# kubectl apply -f stat-3.yaml ; kubectl get po -w
pod/stat-3 created
NAME              READY   STATUS              RESTARTS      AGE
liveness-1        1/1     Running             0             19h
liveness-3        1/1     Running             0             18h
periodseconds-1   1/1     Running             0             18h
periodseconds-2   1/1     Running             0             15h
pod-test-1        1/1     Running             0             22h
pod-test-2        1/1     Running             0             22h
pod-test-cx       1/1     Running             1 (35h ago)   4d20h
stat-1            1/1     Running             0             12h
stat-2            1/1     Running             0             5m
stat-3            0/1     ContainerCreating   0             1s
stat-3            0/1     ContainerCreating   0             1s
stat-3            0/1     Running             0             2s
stat-3            0/1     Running             0             11s
stat-3            1/1     Running             0             11s

容器重啟策略：　　restartPolicy

restartPolicy 應用於 Pod 中的應用容器和常規的 Init 容器。 Sidecar 容器忽略 Pod 級別的 restartPolicy 欄位：在 Kubernetes 中，Sidecar 被定義為 initContainers 內的一個條目，其容器級別的 restartPolicy 被設定為 Always。對於因錯誤而退出的 Init 容器，如果 Pod 級別 restartPolicy 為 OnFailure 或 Always，則 kubelet 會重新啟動 Init 容器。

Always：只要容器終止就自動重啟容器。
OnFailure：只有在容器錯誤退出（退出狀態非零）時才重新啟動容器。
Never：不會自動重啟已終止的容器。

混合使用

[root@master pod]# cat readinessProbe-3.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: lrso
  labels: 
    app: exec-1
spec:
  containers:
  - name: liveness
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - name: web
      containerPort: 80
    readinessProbe:
      initialDelaySeconds: 10
      exec:
        command:
        - /bin/sh
        - "-c"
        - "ls /usr/share/nginx/html/index.html"
      successThreshold: 1
      periodSeconds: 3
    livenessProbe:
      initialDelaySeconds: 15
      httpGet:
        port: web
        path: /index.html
    startupProbe:
      httpGet:
        path: /index.html
        port: web
      failureThreshold: 3 # 失敗的次數
      periodSeconds: 10 # 探測間隔
 
---
apiVersion: v1
kind: Service
metadata:
  name: riodseconds-exec-1
spec:
  selector:
    app: exec-1
  ports:
  - name: web
    protocol: TCP
    targetPort: web
    port: 80
[root@master pod]# kubectl apply -f readinessProbe-3.yaml ; kubectl get po -w
pod/lrso created
service/riodseconds-exec-1 created
NAME              READY   STATUS              RESTARTS      AGE
liveness-1        1/1     Running             0             20h
liveness-3        1/1     Running             0             19h
lrso              0/1     ContainerCreating   0             1s
periodseconds-1   1/1     Running             0             19h
periodseconds-2   1/1     Running             0             16h
pod-test-1        1/1     Running             0             22h
pod-test-2        1/1     Running             0             22h
pod-test-cx       1/1     Running             1 (35h ago)   4d21h
stat-1            1/1     Running             0             12h
stat-2            1/1     Running             0             35m
stat-3            1/1     Running             0             30m
lrso              0/1     ContainerCreating   0             1s
lrso              0/1     Running             0             2s
lrso              0/1     Running             0             11s
lrso              1/1     Running             0             11s

Ruby 探針的基本實現原理
2019-02-16
高階語言實現的幾個點
2019-04-13
Kubernetes-15：一文詳解Pod、Node排程規則(親和性、汙點、容忍、固定節點)
2020-09-03
Laravel集合探學系列——高階訊息傳遞實現（二）
2018-08-08
Laravel
淺入Kubernetes(10)：控制節點的部署，選擇器、親和性、汙點
2021-04-23
Web前端------JS高階繼承的實現方式
2018-07-07
Web前端JS繼承
Nginx 高階篇（三）負載均衡的實現
2020-03-18
Nginx負載
k8s-Pod汙點與容忍
2021-11-25
K8S
淺探VUE的MVVM模式實現
2018-11-16
VueMVVM模式
PHP外掛系統的實現(三):實現探測器
2021-06-15
PHP
dubbo 協議的 K8s pod 存活探針配置
2021-12-06
協議K8S
Pod就緒性探測
2019-09-29
Kubenetes裡pod和service繫結的實現方式
2018-10-06
使用 MaxScale 實現資料庫的高可用性和彈性
2024-04-13
資料庫
Kubernetes Pod配置：從基礎到高階實戰技巧
2024-04-12
在K8S中，Pod 如何實現對節點的資源控制?
2024-10-15
K8S
vue + axios 實現分頁引數傳遞，高階搜尋功能實現
2020-12-18
VueiOS
掌握SpringBoot-2.3的容器探針：實戰篇
2020-09-18
Spring Boot
JuiceFS CSI：Mount Pod 的平滑升級及其實現原理
2024-11-01
UI
Kubernetes Pod排程：從基礎到高階實戰技巧
2024-05-20
實現個性化營銷的CRM有哪幾點
2022-11-09
Nginx實現請求的負載均衡 + keepalived實現Nginx的高可用
2019-03-06
Nginx負載
如何利用 Seaborn 實現高階統計圖表
2024-05-20
巧用漸變實現高階感拉滿的背景光動畫
2021-11-24
動畫
使用keepalived實現nginx的高可用
2019-02-17
Nginx
用程式碼探討KVC/KVO的實現原理
2019-02-27
用程式碼探討 KVC/KVO 的實現原理
2018-04-05
PoD-Tiny——實現零信任交易的最簡協議
2019-07-24
協議
Laravel 動態屬性的實現
2018-04-13
Laravel
面試高頻點-Promise解析和實現
2018-12-12
面試Promise
實現高可用的兩種方案與實戰
2018-08-30
python實現密碼破解
2020-11-10
Python密碼
密碼學課程設計 - 混合密碼的實現
2020-12-05
密碼學
Go 如何實現 PHP 的密碼加密解密
2019-04-16
GoPHP密碼加密解密
觀點：實現CQRS分離不如實現一致性分離 - @jroper
2019-01-16
iOS 實現自動登入（從低階做法到高階做法）
2018-08-02
iOS
商品屬性的選擇功能的實現
2018-04-24
Vue原始碼探祕（五）（_render 函式的實現）
2020-03-23
Vue原始碼函式

pod 的高階實現汙點親密性探針的實現

相關文章