Skip to content

README.adoc

Latest commit

 

History

History
141 lines (103 loc) · 3.74 KB

README.adoc

File metadata and controls

141 lines (103 loc) · 3.74 KB

Stateless Service: Executable Guide

We recommend using a Minikube installation for this example. For details, please refer to the installation instructions.

We assume you use the default 2GB memory for your Minikube VM to play with memory limits. You can adjust this value with the --memory flag. ; To access the PersistentVolume used in this demo, mount a local directory logs/ into the Minikube VM with the following command:

minikube start --mount --mount-string="$(pwd)/logs:/tmp/example" --memory 2G

The directory is now available within the Minikube VM at the path /tmp/example.

For this example, we will use the random number service, a simple REST service that returns just a random number. This service is in the image k8spatterns/random-generator:1.0 from Docker Hub.

Create a ReplicaSet

First, create a ReplicaSet for your stateless service:

kubectl apply -f https://k8spatterns.com/StatelessService/replicaset.yml

Verify the ReplicaSet is working correctly:

kubectl get rs
NAME               DESIRED   CURRENT   READY   AGE
random-generator   3         3         3       17s
kubectl describe rs random-generator
Name:         random-generator
Namespace:    default
Selector:     app=random-generator
Labels:       <none>
Annotations:  <none>
Replicas:     3 current / 3 desired
Pods Status:  3 Running / 0 Waiting / 0 Succeeded / 0 Failed
....

You should see that all three Pods are coming up as expected.

Self-Healing

Let’s play with the ReplicaSet and check whether the self-healing works. For this, let’s kill a random Pod:

kubectl delete $(kubectl get pods -l app=random-generator -o name | head -1)

Check whether a new Pod is created so that the declared state of 3 replicas is fulfilled again)

kubectl get pods
NAME                     READY   STATUS    RESTARTS   AGE
random-generator-8r727   1/1     Running   0          5s
random-generator-qkv2c   1/1     Running   0          2m1s
random-generator-smtsp   1/1     Running   0          2m1s

Access via a Service

To access the ReplicaSet from within the cluster, we create a Service of type clusterIP:

kubectl apply -f https://k8spatterns.com/StatelessService/service.yml

Let’s test the service by calling the Pod from within the cluster:

kubectl run -itq --rm --image=k8spatterns/curl-jq \
    --restart=Never curl -- http://random-generator:8080

Can you see how different Pods are hit when you repeat this command multiple times?

Using a Persistent Volume

Create a PersistentVolume and a PersistentVolumeClaim:

kubectl apply -f https://k8spatterns.com/StatelessService/pv-and-pvc.yml

Update the ReplicaSet to use the PersistentVolumeClaim:

kubectl apply -f https://k8spatterns.com/StatelessService/replicaset-with-pv.yml

Delete all Pods so that they are recreated with the new configuration:

kubectl delete pod -l app=random-generator

Call several times again our service with

kubectl run -itq --rm --image=k8spatterns/curl-jq \
    --restart=Never curl -- http://random-generator:8080

Now let’s check your local directory logs/ and see how it gets populated with log files from the three Pods