

Installing the API K8s Operator¶

Follow the instructions given below to install and set up the API K8s Operator in your Kubernetes environment.

Before you begin¶

Install kubectl.
Setup a Kubernetes cluster v1.14 or above.
- Minimum CPU : 2vCPU
- Minimum Memory : 2GB
Alternatively, you can run Kubernetes locally via Minikube.

Install the API K8s Operator in Standalone mode¶

Follow the steps given below to install the operator in standalone mode.

Install the API Operator:

kubectl apply -f https://github.com/wso2/k8s-api-operator/releases/download/2.0.2/api-operator-configs.yaml

Verify the installation:

kubectl get pods

Output:
NAME                            READY   STATUS              RESTARTS   AGE
api-operator-776c79b5f6-c9n9n   1/1     Running             0          56s

Install K8s API Operator using OperatorHub.io¶

For information, see the installation instructions in the OperatorHub.io.

Install API K8s Operator in Cluster mode¶

By default, the K8s API operator is configured to watch the deployed namespace. That is, if the API Operator runs on the default namespace, it can only watch the custom resources that are deployed in the default namespace. By deploying in cluster mode, the API Operator is able to watch all the custom resources in all the namespaces.

Deploy the ClusterRole:

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
   name: api-operator
rules:
   - apiGroups:
        - ""
     resources:
        - pods
        - services
        - endpoints
        - persistentvolumeclaims
        - events
        - configmaps
        - secrets
        - ingresses
     verbs:
        - '*'
   - apiGroups:
        - apps
     resources:
        - deployments
        - daemonsets
        - replicasets
        - statefulsets
        - ingresses
     verbs:
        - '*'
   - apiGroups:
        - monitoring.coreos.com
     resources:
        - servicemonitors
     verbs:
        - get
        - create
   - apiGroups:
        - apps
     resourceNames:
        - api-operator
     resources:
        - deployments/finalizers
     verbs:
        - update
   - apiGroups:
        - wso2.com
     resources:
        - '*'
        - ratelimitings
        - targetendpoints
        - securities
        - integrations
     verbs:
        - '*'
   - apiGroups:
        - batch
     resources:
        - '*'
     verbs:
        - '*'
   - apiGroups:
        - autoscaling
     resources:
        - '*'
     verbs:
        - '*'
   - apiGroups:
        - extensions
        - networking.k8s.io
     resources:
        - ingresses
        - ingress
        - ingressclasses
     verbs:
        - '*'
   - apiGroups:
        - serving.knative.dev
     resources:
        - '*'
     verbs:
        - get
        - list
        - create
        - update
        - delete
        - patch
        - watch
   - apiGroups:
        - route.openshift.io
     resources:
        - '*'
     verbs:
        - get
        - list
        - create
        - update
        - delete
        - patch
        - watch
   - apiGroups:
        - networking.istio.io
     resources:
        - '*'
     verbs:
        - get
        - list
        - create
        - update
        - delete
        - patch
        - watch

Deploy the ClusterRoleBinding:

Be sure to replace <OPERATOR_DEPLOYED_NAMESPACE> with the correct value.

kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
   name: api-operator
subjects:
   - kind: ServiceAccount
     name: api-operator
      # Replace this with the namespace the operator is deployed in.
     namespace: <OPERATOR_DEPLOYED_NAMESPACE>
roleRef:
   kind: ClusterRole
   name: api-operator
   apiGroup: rbac.authorization.k8s.io

Update API operator to watch the cluster:

If the operator is already running in the Kubernetes cluster, use the following command to set the WATCH_CLUSTER_LEVEL configuration to true in the API operator deployment.
```
kubectl set env deployment/api-operator WATCH_CLUSTER_LEVEL=true
```

What's Next¶

Top