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:
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:
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:
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:
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:
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:
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:
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:
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:
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