learning kubernetes by minikube,

作者: caccq | 来源:发表于2018-04-16 16:31 被阅读0次

playing with probes.

Goal

Add HTTP health checks to the helloworld deployment
Simulate a failing deployment that fails a readiness probe
Simulate a failing deployment that fails a liveness probe

exist helloworld deployment yaml file as below

apiVersion: apps/v1beta1
kind: Deployment
metadata:
  name: helloworld-deployment
spec:
  selector:
    matchLabels:
      app: helloworld
  replicas: 1 # tells deployment to run 1 pods matching the template
  template: # create pods using pod definition in this template
    metadata:
      labels:
        app: helloworld
    spec:
      containers:
      - name: helloworld
        image: karthequian/helloworld:latest
        ports:
        - containerPort: 80

A readiness probe is used to know when a container is ready to start accepting traffic.

readinessProbe:
  # length of time to wait for a pod to initialize
  # after pod startup, before applying health checking
  initialDelaySeconds: 10
  # Amount of time to wait before timing out
  initialDelaySeconds: 1
  # Probe for http
  httpGet:
    # Path to probe
    path: /
    # Port to probe
    port: 80

A liveness probe is used to know when a container might need to be restarted.

livenessProbe:
  # length of time to wait for a pod to initialize
  # after pod startup, before applying health checking
  initialDelaySeconds: 10
  # Amount of time to wait before timing out
  timeoutSeconds: 1
  # Probe for http
  httpGet:
    # Path to probe
    path: /
    # Port to probe
    port: 80

after add the liveness and readiness probe, the yaml file become as
helloworld-with-probes.yaml

apiVersion: apps/v1beta1
kind: Deployment
metadata:
  name: helloworld-deployment-with-probe
spec:
  selector:
    matchLabels:
      app: helloworld
  replicas: 1 # tells deployment to run 1 pods matching the template
  template: # create pods using pod definition in this template
    metadata:
      labels:
        app: helloworld
    spec:
      containers:
      - name: helloworld
        image: karthequian/helloworld:latest
        ports:
        - containerPort: 80
        readinessProbe:
          # length of time to wait for a pod to initialize
          # after pod startup, before applying health checking
          initialDelaySeconds: 10
          # Amount of time to wait before timing out
          initialDelaySeconds: 1
          # Probe for http
          httpGet:
            # Path to probe
            path: /
            # Port to probe
            port: 80
        livenessProbe:
          # length of time to wait for a pod to initialize
          # after pod startup, before applying health checking
          initialDelaySeconds: 10
          # Amount of time to wait before timing out
          timeoutSeconds: 1
          # Probe for http
          httpGet:
            # Path to probe
            path: /
            # Port to probe
            port: 80

localhost:~ xunyang$ kubectl create -f helloworld-with-probes.yaml 
deployment.apps "helloworld-deployment-with-probe" created
localhost:~ xunyang$ kubectl get deployment
NAME                               DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
helloworld-deployment-with-probe   1         1         1            1           9m
localhost:~ xunyang$ kubectl describe pod helloworld-deployment-with-probe-7f78698c95-whltr
Name:           helloworld-deployment-with-probe-7f78698c95-whltr
Namespace:      default
Node:           minikube/10.0.2.15
Start Time:     Mon, 16 Apr 2018 15:44:07 +0800
Labels:         app=helloworld
                pod-template-hash=3934254751
Annotations:    <none>
Status:         Running
IP:             172.17.0.4
Controlled By:  ReplicaSet/helloworld-deployment-with-probe-7f78698c95
Containers:
  helloworld:
    Container ID:   docker://f27e53a391706040d16409054b9f4792b47e5e6a5ef5749fc839000bf11345ee
    Image:          karthequian/helloworld:latest
    Image ID:       docker-pullable://karthequian/helloworld@sha256:165f87263f1775f0bf91022b48a51265357ba9f36bc3882f5ecdefc7f8ef8f6d
    Port:           80/TCP
    Host Port:      0/TCP
    State:          Running
      Started:      Mon, 16 Apr 2018 15:44:11 +0800
    Ready:          True
    Restart Count:  0
    Liveness:       http-get http://:80/ delay=10s timeout=1s period=10s #success=1 #failure=3
    Readiness:      http-get http://:80/ delay=1s timeout=1s period=10s #success=1 #failure=3
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-64pmj (ro)
Conditions:
  Type           Status
  Initialized    True 
  Ready          True 
  PodScheduled   True 
Volumes:
  default-token-64pmj:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-64pmj
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type    Reason                 Age   From               Message
  ----    ------                 ----  ----               -------
  Normal  Scheduled              23m   default-scheduler  Successfully assigned helloworld-deployment-with-probe-7f78698c95-whltr to minikube
  Normal  SuccessfulMountVolume  23m   kubelet, minikube  MountVolume.SetUp succeeded for volume "default-token-64pmj"
  Normal  Pulling                23m   kubelet, minikube  pulling image "karthequian/helloworld:latest"
  Normal  Pulled                 23m   kubelet, minikube  Successfully pulled image "karthequian/helloworld:latest"
  Normal  Created                23m   kubelet, minikube  Created container
  Normal  Started                23m   kubelet, minikube  Started container

Simulate a failing deployment that fails a readiness probe

We will now try to simulate a bad helloworld pod that fails a readiness probe. Run a readiness check on port 90 to simulate a failing scenario.

helloworld-with-bad-readiness-probe.yaml

apiVersion: apps/v1beta1
kind: Deployment
metadata:
  name: helloworld-deployment-with-bad-readiness-probe
spec:
  selector:
    matchLabels:
      app: helloworld
  replicas: 1 # tells deployment to run 1 pods matching the template
  template: # create pods using pod definition in this template
    metadata:
      labels:
        app: helloworld
    spec:
      containers:
      - name: helloworld
        image: karthequian/helloworld:latest
        ports:
        - containerPort: 80
        readinessProbe:
          # length of time to wait for a pod to initialize
          # after pod startup, before applying health checking
          initialDelaySeconds: 10
          # Amount of time to wait before timing out
          initialDelaySeconds: 1
          # Probe for http
          httpGet:
            # Path to probe
            path: /
            # Port to probe
            port: 90

localhost:~ xunyang$ kubectl create -f helloworld-with-bad-readiness-probe.yaml 
deployment.apps "helloworld-deployment-with-bad-readiness-probe" created
localhost:~ xunyang$ kubectl get deployment
NAME                                             DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
helloworld-deployment-with-bad-readiness-probe   1         1         1            0           9m
helloworld-deployment-with-probe                 1         1         1            1           12m
localhost:~ xunyang$ kubectl get pods
NAME                                                             READY     STATUS    RESTARTS   AGE
helloworld-deployment-with-bad-readiness-probe-9447b7b49-p2qwm   0/1       Running   0          10m
helloworld-deployment-with-probe-7f78698c95-whltr                1/1       Running   0          13m
localhost:~ xunyang$ kubectl describe pod helloworld-deployment-with-bad-readiness-probe-9447b7b49-p2qwm
Name:           helloworld-deployment-with-bad-readiness-probe-9447b7b49-p2qwm
Namespace:      default
Node:           minikube/10.0.2.15
Start Time:     Mon, 16 Apr 2018 15:47:15 +0800
Labels:         app=helloworld
                pod-template-hash=500363605
Annotations:    <none>
Status:         Running
IP:             172.17.0.5
Controlled By:  ReplicaSet/helloworld-deployment-with-bad-readiness-probe-9447b7b49
Containers:
  helloworld:
    Container ID:   docker://a2fdd491e851524fa5f24ea7d20e0306004561a43c7e734b4992ae3da71859a2
    Image:          karthequian/helloworld:latest
    Image ID:       docker-pullable://karthequian/helloworld@sha256:165f87263f1775f0bf91022b48a51265357ba9f36bc3882f5ecdefc7f8ef8f6d
    Port:           80/TCP
    Host Port:      0/TCP
    State:          Running
      Started:      Mon, 16 Apr 2018 15:47:19 +0800
    Ready:          False
    Restart Count:  0
    Readiness:      http-get http://:90/ delay=1s timeout=1s period=10s #success=1 #failure=3
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-64pmj (ro)
Conditions:
  Type           Status
  Initialized    True 
  Ready          False 
  PodScheduled   True 
Volumes:
  default-token-64pmj:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-64pmj
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason                 Age                From               Message
  ----     ------                 ----               ----               -------
  Normal   Scheduled              12m                default-scheduler  Successfully assigned helloworld-deployment-with-bad-readiness-probe-9447b7b49-p2qwm to minikube
  Normal   SuccessfulMountVolume  12m                kubelet, minikube  MountVolume.SetUp succeeded for volume "default-token-64pmj"
  Normal   Pulling                12m                kubelet, minikube  pulling image "karthequian/helloworld:latest"
  Normal   Pulled                 12m                kubelet, minikube  Successfully pulled image "karthequian/helloworld:latest"
  Normal   Created                12m                kubelet, minikube  Created container
  Normal   Started                12m                kubelet, minikube  Started container
  Warning  Unhealthy              9m (x20 over 12m)  kubelet, minikube  Readiness probe failed: Get http://172.17.0.5:90/: dial tcp 172.17.0.5:90: getsockopt: connection refused

as the pod is not ready, the pod check will wait till the pod is ready.

Simulate a failing deployment that fails a liveness probe

we will simulate a bad helloworld pod that fails a liveness probe. Run a liveness check on port 90 to simulate a failing scenario.

helloworld-with-bad-liveness-probe.yaml

apiVersion: apps/v1beta1
kind: Deployment
metadata:
  name: helloworld-deployment-with-bad-liveness-probe
spec:
  selector:
    matchLabels:
      app: helloworld
  replicas: 1 # tells deployment to run 1 pods matching the template
  template: # create pods using pod definition in this template
    metadata:
      labels:
        app: helloworld
    spec:
      containers:
      - name: helloworld
        image: karthequian/helloworld:latest
        ports:
        - containerPort: 80
        livenessProbe:
          # length of time to wait for a pod to initialize
          # after pod startup, before applying health checking
          initialDelaySeconds: 10
          # How often (in seconds) to perform the probe.
          periodSeconds: 5
          # Amount of time to wait before timing out
          timeoutSeconds: 1
          # Kubernetes will try failureThreshold times before giving up and restarting the Pod
          failureThreshold: 2
          # Probe for http
          httpGet:
            # Path to probe
            path: /
            # Port to probe
            port: 90

localhost:~ xunyang$ kubectl create -f helloworld-with-bad-liveness-probe.yaml
deployment.apps "helloworld-deployment-with-bad-liveness-probe" created
ocalhost:~ xunyang$ kubectl get pods
NAME                                                             READY     STATUS             RESTARTS   AGE
helloworld-deployment-with-bad-liveness-probe-7fbddcfd4d-8k6t8   0/1       CrashLoopBackOff   6          13m
helloworld-deployment-with-bad-readiness-probe-9447b7b49-p2qwm   0/1       Running            0          22m
helloworld-deployment-with-probe-7f78698c95-whltr                1/1       Running            0          25m
localhost:~ xunyang$ kubectl describe pod helloworld-deployment-with-bad-liveness-probe-7fbddcfd4d-8k6t8
Name:           helloworld-deployment-with-bad-liveness-probe-7fbddcfd4d-8k6t8
Namespace:      default
Node:           minikube/10.0.2.15
Start Time:     Mon, 16 Apr 2018 15:55:35 +0800
Labels:         app=helloworld
                pod-template-hash=3968879808
Annotations:    <none>
Status:         Running
IP:             172.17.0.6
Controlled By:  ReplicaSet/helloworld-deployment-with-bad-liveness-probe-7fbddcfd4d
Containers:
  helloworld:
    Container ID:   docker://32d999bbc20c8c00dc0e19a51a351b82a7cbf763f38715a3b0791ec86e7b3f0f
    Image:          karthequian/helloworld:latest
    Image ID:       docker-pullable://karthequian/helloworld@sha256:165f87263f1775f0bf91022b48a51265357ba9f36bc3882f5ecdefc7f8ef8f6d
    Port:           80/TCP
    Host Port:      0/TCP
    State:          Waiting
      Reason:       CrashLoopBackOff
    Last State:     Terminated
      Reason:       Completed
      Exit Code:    0
      Started:      Mon, 16 Apr 2018 15:59:12 +0800
      Finished:     Mon, 16 Apr 2018 15:59:24 +0800
    Ready:          False
    Restart Count:  6
    Liveness:       http-get http://:90/ delay=10s timeout=1s period=5s #success=1 #failure=2
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-64pmj (ro)
Conditions:
  Type           Status
  Initialized    True 
  Ready          False 
  PodScheduled   True 
Volumes:
  default-token-64pmj:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-64pmj
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason                 Age                From               Message
  ----     ------                 ----               ----               -------
  Normal   Scheduled              14m                default-scheduler  Successfully assigned helloworld-deployment-with-bad-liveness-probe-7fbddcfd4d-8k6t8 to minikube
  Normal   SuccessfulMountVolume  14m                kubelet, minikube  MountVolume.SetUp succeeded for volume "default-token-64pmj"
  Normal   Pulling                13m (x4 over 14m)  kubelet, minikube  pulling image "karthequian/helloworld:latest"
  Normal   Killing                13m (x3 over 14m)  kubelet, minikube  Killing container with id docker://helloworld:Container failed liveness probe.. Container will be killed and recreated.
  Normal   Pulled                 13m (x4 over 14m)  kubelet, minikube  Successfully pulled image "karthequian/helloworld:latest"
  Normal   Created                13m (x4 over 14m)  kubelet, minikube  Created container
  Normal   Started                13m (x4 over 14m)  kubelet, minikube  Started container
  Warning  Unhealthy              13m (x5 over 14m)  kubelet, minikube  Liveness probe failed: Get http://172.17.0.6:90/: dial tcp 172.17.0.6:90: getsockopt: connection refused

as the liveness check failed, the pod will continue recreate even the pod is ready.

localhost:~ xunyang$ kubectl get pods
NAME                                                             READY     STATUS    RESTARTS   AGE
helloworld-deployment-with-bad-liveness-probe-7fbddcfd4d-8k6t8   1/1       Running   5          11m
helloworld-deployment-with-bad-readiness-probe-9447b7b49-p2qwm   0/1       Running   0          19m
helloworld-deployment-with-probe-7f78698c95-whltr                1/1       Running   0          22m
localhost:~ xunyang$ kubectl get pod
NAME                                                             READY     STATUS             RESTARTS   AGE
helloworld-deployment-with-bad-liveness-probe-7fbddcfd4d-8k6t8   0/1       CrashLoopBackOff   11         29m
helloworld-deployment-with-bad-readiness-probe-9447b7b49-p2qwm   0/1       Running            0          37m
helloworld-deployment-with-probe-7f78698c95-whltr                1/1       Running            0          40m
localhost:~ xunyang$ kubectl get pod

when check deployment status, the both wrong probe shows not available.

localhost:~ xunyang$ kubectl get deployment
NAME                                             DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
helloworld-deployment-with-bad-liveness-probe    1         1         1            0           16m
helloworld-deployment-with-bad-readiness-probe   1         1         1            0           24m
helloworld-deployment-with-probe                 1         1         1            1           27m

learning kubernetes by minikube,

playing with probes.

Goal

Simulate a failing deployment that fails a readiness probe

Simulate a failing deployment that fails a liveness probe

clear data

相关文章

网友评论

延伸阅读

深度阅读

栏目导航

热点阅读

我爱编程