Starting and Stopping OIG using kubectl

Before You Begin

This tutorial shows you how to perform basic operations with kubectl to stop and start the Oracle Identity Governance (OIG) Managed Servers and AdminServer on Oracle Cloud Infrastructure (OCI).

Background

When OIG is deployed in OCI via OCI MarketPlace, it is deployed in a Kubernetes (K8S) cluster using the Oracle Kubernetes Engine (OKE).

Whilst WebLogic Console and Oracle Enterprise Manager Console can be used for normal monitoring and administration operations, it cannot be used to stop and start the OIG AdminServer, OIG Managed Servers or SOA Managed Server. The lifecycle of the OIG WebLogic Servers in K8S is managed using the Oracle WebLogic Server Kubernetes Operator and all start and stop operations must be performed using kubectl commands.

What Do You Need?

To have completed the tutorial Deploying Oracle Identity Governance on Oracle Cloud Infrastructure
An SSH connection to the bastion host
The external IP address of the load balancer service

View the OIG Servers using kubectl

The default OIG deployment using OCI MarketPlace starts the AdminServer (AdminServer), one OIG Managed Server (oim_server1), and one SOA Managed Server (soa_server1).

The deployment also creates, but doesn't start, four extra OIG Managed Servers (oim-server2 - oig-server5) and three more SOA Managed Servers (soa_server2 - soa_server4).

All these servers are visible in the WebLogic Server Console (http://<external-ip>/console) by navigating to Domain Structure > OIGcluster > Environment > Servers.

To view the running servers using kubectl, perform the following:

Connect to the bastion host using the cluster.key file:
```
$ ssh -i cluster.key opc@<bastion_ip>
```

On the bastion host, run the following command to view the pods associated with the OIG servers:

$ kubectl get pods -n oigcluster

The output should look similar to the following:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          46m
oigcluster-create-fmw-infra-sample-domain-job-s9vqm   0/1     Completed   0          68m
oigcluster-oim-server1                                1/1     Running     0          35m
oigcluster-soa-server1                                1/1     Running     0          41m

Here you can see the pods that are currently running (oigcluster-adminserver, oigcluster-oim-server1, oigcluster-soa-server1). The READY status 1/1 indicates the associated servers are also up and running.

Starting OIG Managed Servers

The number of OIG Managed Servers running is dependent on the replicas parameter configured for the cluster. To start more OIG Managed Servers perform the following steps:

Run the following kubectl command to edit the oigcluster:
```
$ kubectl edit domain oigcluster -n oigcluster
```
Note: This opens an edit session for the domain where parameters can be changed using standard vi commands.

In the edit session search for "clusterName: oim_cluster" and look for the replicas parameter. By default the replicas parameter is set to "1" hence one OIG Managed Server is started (oim_server1):

- clusterName: oim_cluster
      replicas: 1
      serverPod:
        affinity:
          podAntiAffinity:
            preferredDuringSchedulingIgnoredDuringExecution:
            - podAffinityTerm:
                labelSelector:
                  matchExpressions:
                  - key: weblogic.clusterName
                    operator: In
                    values:
                    - $(CLUSTER_NAME)
                topologyKey: kubernetes.io/hostname
              weight: 100

To start more OIG Managed Servers, increase the replicas value as desired. In the example below, two more managed servers will be started by setting replicas to "3":

- clusterName: oim_cluster
      replicas: 3
      serverPod:
        affinity:
          podAntiAffinity:
            preferredDuringSchedulingIgnoredDuringExecution:
            - podAffinityTerm:
                labelSelector:
                  matchExpressions:
                  - key: weblogic.clusterName
                    operator: In
                    values:
                    - $(CLUSTER_NAME)
                topologyKey: kubernetes.io/hostname
              weight: 100

Note: If scaling up and starting more servers, ensure you have a high powered compute with enough RAM and CPU in order to prevent problems with pod restart.

Save the file and exit (:wq!).

The output will look similar to the following:
```
domain.weblogic.oracle/oigcluster edited
```

Run the following kubectl command to view the pods:

kubectl get pods -n oigcluster

The output will look similar to the following:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          49m
oigcluster-create-fmw-infra-sample-domain-job-s9vqm   0/1     Completed   0          71m
oigcluster-oim-server1                                1/1     Running     0          38m
oigcluster-oim-server2                                0/1     Running     0          19s
oigcluster-oim-server3                                0/1     Running     0          19s
oigcluster-soa-server1                                1/1     Running     0          44m

Two new pods (oigcluster-oim-server2 and oigcluster-oim-server3) are started, but currently have a READY status of 0/1. This means oim_server2 and oim_server3 are not currently running but are in the process of starting. The servers will take several minutes to start so keep executing the command until READY shows 1/1:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          76m
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-oim-server1                                1/1     Running     0          66m
oigcluster-oim-server2                                1/1     Running     0          27m
oigcluster-oim-server3                                1/1     Running     0          27m
oigcluster-soa-server1                                1/1     Running     0          71m

Note: To check what is happening during server startup when READY is 0/1, run the following command to view the log of the pod that is starting:

kubectl logs oigcluster-oim-server3 -n oigcluster

To start more SOA Managed servers, run the following kubectl command to edit the oigcluster:
```
$ kubectl edit domain oigcluster -n oigcluster
```

Search for "clusterName: soa_cluster" and look for the replicas parameter. By default the replicas parameter is set to "1" hence one SOA Managed server is started (soa_server1):

 - clusterName: soa_cluster
      replicas: 1
      serverPod:
        affinity:
          podAntiAffinity:
            preferredDuringSchedulingIgnoredDuringExecution:
            - podAffinityTerm:
                labelSelector:
                  matchExpressions:
                  - key: weblogic.clusterName
                    operator: In
                    values:
                    - $(CLUSTER_NAME)
                topologyKey: kubernetes.io/hostname
              weight: 100

To start more SOA Managed Servers, increase the replicas value as desired. In the example below, one more SOA Managed server will be started by setting replicas to "2":

- clusterName: soa_cluster
      replicas: 2
      serverPod:
        affinity:
          podAntiAffinity:
            preferredDuringSchedulingIgnoredDuringExecution:
            - podAffinityTerm:
                labelSelector:
                  matchExpressions:
                  - key: weblogic.clusterName
                    operator: In
                    values:
                    - $(CLUSTER_NAME)
                topologyKey: kubernetes.io/hostname
              weight: 100

Save the file and exit (:wq!).

Run the following kubectl command to view the pods:

kubectl get pods -n oigcluster

The output will look similar to the following:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          76m
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-oim-server1                                1/1     Running     0          66m
oigcluster-oim-server2                                1/1     Running     0          27m
oigcluster-oim-server3                                1/1     Running     0          27m
oigcluster-soa-server1                                1/1     Running     0          71m
oigcluster-soa-server2                                0/1     Running     0          2m3s

A new pod (oigcluster-soa-server2) is started, but currently has a READY status of 0/1. The server will take several minutes to start so keep executing the command until READY shows 1/1:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          76m
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-oim-server1                                1/1     Running     0          66m
oigcluster-oim-server2                                1/1     Running     0          29m
oigcluster-oim-server3                                1/1     Running     0          29m
oigcluster-soa-server1                                1/1     Running     0          71m
oigcluster-soa-server2                                1/1     Running     0          4m50s

Stopping OIG Managed Servers

As mentioned in the previous section, the number of OIG Managed Servers running is dependent on the replicas parameter configured for the cluster. To stop one or more OIG Managed Servers, perform the following:

Run the following kubectl command to edit the OIGcluster:
```
$ kubectl edit domain oigcluster -n oigcluster
```

In the edit session search for "clusterName: oim_cluster" and look for the replicas parameter. Assuming the replicas parameter was increased in the previous section, the current value is set to "3" hence three OIG Managed Servers are started (oim_server1 - oim_server3):

- clusterName: oim_cluster
      replicas: 3
      serverPod:
        affinity:
          podAntiAffinity:
            preferredDuringSchedulingIgnoredDuringExecution:
            - podAffinityTerm:
                labelSelector:
                  matchExpressions:
                  - key: weblogic.clusterName
                    operator: In
                    values:
                    - $(CLUSTER_NAME)
                topologyKey: kubernetes.io/hostname
              weight: 100

To stop OIG Managed Servers, decrease the replicas value as desired. In the example below, we will stop two managed servers by setting replicas to "1":

- clusterName: oim_cluster
      replicas: 1
      serverPod:
        affinity:
          podAntiAffinity:
            preferredDuringSchedulingIgnoredDuringExecution:
            - podAffinityTerm:
                labelSelector:
                  matchExpressions:
                  - key: weblogic.clusterName
                    operator: In
                    values:
                    - $(CLUSTER_NAME)
                topologyKey: kubernetes.io/hostname
              weight: 100

Save the file and exit (:wq!).

Run the following kubectl command to view the pods:

kubectl get pods -n oigcluster

The output will look similar to the following:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          76m
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-oim-server1                                1/1     Running     0          66m
oigcluster-oim-server2                                1/1     Terminating 0          29m
oigcluster-oim-server3                                1/1     Terminating 0          29m
oigcluster-soa-server1                                1/1     Running     0          71m
oigcluster-soa-server2                                1/1     Running     0          4m50s

Two pods now have a STATUS of Terminating (oigcluster-oim-server2 and oigcluster-oim-server3). The servers will take a minute or two to stop, so keep executing the command until the pods have disappeared:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          76m
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-oim-server1                                1/1     Running     0          66m
oigcluster-soa-server1                                1/1     Running     0          71m
oigcluster-soa-server2                                1/1     Running     0          4m50s

To stop SOA Managed Servers run the following kubectl command to edit the oigcluster:
```
$ kubectl edit domain oigcluster -n oigcluster
```

In the edit session search for "clusterName: soa_cluster" and look for the replicas parameter. Assuming the replicas parameter was increased in the previous section, the current value is set to "2" hence two SOA Managed Servers are started (soa_server1 and soa_server2):

- clusterName: soa_cluster
      replicas: 2
      serverPod:
        affinity:
          podAntiAffinity:
            preferredDuringSchedulingIgnoredDuringExecution:
            - podAffinityTerm:
                labelSelector:
                  matchExpressions:
                  - key: weblogic.clusterName
                    operator: In
                    values:
                    - $(CLUSTER_NAME)
                topologyKey: kubernetes.io/hostname
              weight: 100

To stop SOA Managed Servers, decrease the replicas value as desired. In the example below, we will stop one managed server by setting replicas to "1":

- clusterName: soa_cluster
      replicas: 1
      serverPod:
        affinity:
          podAntiAffinity:
            preferredDuringSchedulingIgnoredDuringExecution:
            - podAffinityTerm:
                labelSelector:
                  matchExpressions:
                  - key: weblogic.clusterName
                    operator: In
                    values:
                    - $(CLUSTER_NAME)
                topologyKey: kubernetes.io/hostname
              weight: 100

Save the file and exit (:wq!).

Run the following kubectl command to view the pods:

kubectl get pods -n oigcluster

The output will look similar to the following:


NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          76m
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-oim-server1                                1/1     Running     0          66m
oigcluster-soa-server1                                1/1     Running     0          71m
oigcluster-soa-server2                                1/1     Terminating 0          4m50s

One pod now has a STATUS of Terminating (oigcluster-soa-server2). The server will take a minute or two to stop, so keep executing the command until the pod has disappeared:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          76m
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-oim-server1                                1/1     Running     0          66m
oigcluster-soa-server1                                1/1     Running     0          71m

Stopping and Starting the AdminServer and Managed Servers

In this section you stop the AdminServer and all the OIG Managed Servers in one operation. After stopping the pods, you then start the pods back up again with a single operation.

Note: At present it is not possible to stop and start the AdminServer on it's own.

Run the following kubectl command to edit the OIGcluster:
```
$ kubectl edit domain oigcluster -n oigcluster
```

In the edit session search for "serverStartPolicy: IF_NEEDED":

...   
    volumes:
    - name: weblogic-domain-storage-volume
      persistentVolumeClaim:
        claimName: oigcluster-oim-pvc
  serverService:
    annotations: {}
    labels: {}
  serverStartPolicy: IF_NEEDED
  webLogicCredentialsSecret:
    name: oigcluster-domain-credentials
status:
...

Change "serverStartPolicy: IF_NEEDED" to "NEVER" as follows:

...   
    volumes:
    - name: weblogic-domain-storage-volume
      persistentVolumeClaim:
        claimName: oigcluster-oim-pvc
  serverService:
    annotations: {}
    labels: {}
  serverStartPolicy: NEVER
  webLogicCredentialsSecret:
    name: oigcluster-domain-credentials
status:

...

Save the file and exit (:wq!).

Run the following kubectl command to view the pods:

kubectl get pods -n oigcluster

The output will look similar to the following:


NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Terminating 0          76m
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-oim-server1                                1/1     Running     0          66m
oigcluster-soa-server1                                1/1     Running     0          71m

The AdminServer pods and Managed Server pods will move to a STATUS of Terminating. After a few minutes, run the command again and the pods should have disappeared:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m

To start the AdminServer and Managed Servers up again, run the following kubectl command to edit the OIGcluster:
```
$ kubectl edit domain oigcluster -n oigcluster
```

In the edit session search for "serverStartPolicy: NEVER":

...   
    volumes:
    - name: weblogic-domain-storage-volume
      persistentVolumeClaim:
        claimName: oigcluster-oim-pvc
  serverService:
    annotations: {}
    labels: {}
  serverStartPolicy: NEVER
  webLogicCredentialsSecret:
    name: oigcluster-domain-credentials
status:

...

Change "serverStartPolicy: NEVER" to "IF_NEEDED" as follows:

...   
 volumes:
    - name: weblogic-domain-storage-volume
      persistentVolumeClaim:
        claimName: OIGcluster-domain-pvc
  serverService:
    annotations: {}
    labels: {}
  serverStartPolicy: IF_NEEDED
  webLogicCredentialsSecret:
    name: OIGcluster-domain-credentials
status:
...

Save the file and exit (:wq!).

Run the following kubectl command to view the pods:

kubectl get pods -n oigcluster

The output will look similar to the following:

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-introspect-domain-job-4sv7z                1/1     Running     0          11s

An oigcluster-introspect-domain-job pod will start. This job will create and start the AdminServer, followed by the OIG Managed Servers pods. This process will take several minutes, so keep executing the command until all the pods are running with READY status 1/1 :

NAME                                                  READY   STATUS      RESTARTS   AGE
oigcluster-adminserver                                1/1     Running     0          76m
oigcluster-create-fmw-infra-sample-domain-job-zfzll   0/1     Completed   0          99m
oigcluster-oim-server1                                1/1     Running     0          66m
oigcluster-soa-server1                                1/1     Running     0          71m

Want to Learn More?

Cloud Infrastructure MarketPlace
Oracle Identity Governance

Feedback

To provide feedback on this tutorial, please contact Identity Management User Assistance.

Starting and Stopping OIG using kubectl

Before You Begin

Background

What Do You Need?

View the OIG Servers using kubectl

Starting OIG Managed Servers

Stopping OIG Managed Servers

Stopping and Starting the AdminServer and Managed Servers

Want to Learn More?

Feedback

Associated Learning Paths