1. You create an AWS IAM role with a trust policy that trusts the cluster’s OpenID Connect (OIDC) provider, scoped to the OpenShift sandboxed containers Operator service account.
2. You attach IAM policies to the role for managing peer pod EC2 instances and creating pod virtual machine (VM) images.
3. When installing the OpenShift sandboxed containers Operator, you provide the IAM role Amazon Resource Name (ARN) as the ROLEARN environment variable in the Operator subscription.
4. The OpenShift sandboxed containers Operator uses the role ARN to create a secret that also references the projected service account token to be consumed by the OpenShift sandboxed containers components.
5. The AWS SDK sends the Kubernetes service account token to AWS STS.
6. AWS STS validates the token against the cluster’s OIDC provider and returns temporary AWS credentials.
7. The OpenShift sandboxed containers components use these temporary credentials to authenticate AWS API requests for peer pod operations.

Procedure

1. Create an osc-namespace.yaml manifest file:

   apiVersion: v1
   kind: Namespace
   metadata:
     name: openshift-sandboxed-containers-operator

   Copy code to clipboard

2. Create the namespace by running the following command:

   $ oc create -f osc-namespace.yaml

   Copy code to clipboard

3. Create an osc-operatorgroup.yaml manifest file:

   apiVersion: operators.coreos.com/v1
   kind: OperatorGroup
   metadata:
     name: sandboxed-containers-operator-group
     namespace: openshift-sandboxed-containers-operator
   spec:
     targetNamespaces:
     - openshift-sandboxed-containers-operator

   Copy code to clipboard

4. Create the Operator group by running the following command:

   $ oc create -f osc-operatorgroup.yaml

   Copy code to clipboard

5. Create an osc-subscription.yaml manifest file:

   apiVersion: operators.coreos.com/v1alpha1
   kind: Subscription
   metadata:
     name: sandboxed-containers-operator
     namespace: openshift-sandboxed-containers-operator
   spec:
     channel: stable
     name: sandboxed-containers-operator
     source: redhat-operators
     sourceNamespace: openshift-marketplace
     startingCSV: sandboxed-containers-operator.v1.13.1

   Copy code to clipboard

6. Optional: If your cluster uses STS authentication, add the following configuration to the osc-subscription.yaml file to set the IAM role ARN:

     # ...
   spec:
     config:
       env:
       - name: ROLEARN
         value: <role_arn>
     installPlanApproval: Manual

   Copy code to clipboard

   <role_arn>

   Specifies the ARN of the IAM role that you created in Section 3.1.1, “Configure an IAM role for STS authentication”. For example, arn:aws:iam::123456789012:role/openshift-sandboxed-containers-role.

   Note

   Manual installation plan approval is recommended for STS deployments. Permission changes might be required before upgrading, and manual subscriptions ensure that the cluster administrator has time to verify the permissions of the next version and take the necessary steps before the upgrade.

7. Create the subscription by running the following command:

   $ oc create -f osc-subscription.yaml

   Copy code to clipboard

Verification

1. Verify that the Operator is correctly installed by running the following command:

   $ oc get csv -n openshift-sandboxed-containers-operator

   Copy code to clipboard

   This command can take several minutes to complete.

2. Watch the installation progress by running the following command:

   $ watch oc get csv -n openshift-sandboxed-containers-operator

   Copy code to clipboard

   NAME                             DISPLAY                                  VERSION         PHASE
   sandboxed-containers-operator.v1.13.1   OpenShift sandboxed containers Operator   1.13.1          Succeeded

   Copy code to clipboard

   The installation is complete when the PHASE column shows Succeeded.

Procedure

1. Log in to your OpenShift Container Platform cluster and retrieve the instance ID:

   $ INSTANCE_ID=$(oc get nodes -l 'node-role.kubernetes.io/worker' \
     -o jsonpath='{.items[0].spec.providerID}' | sed 's#[^ ]*/##g')

   Copy code to clipboard

2. Retrieve the AWS region:

   $ AWS_REGION=$(oc get infrastructure/cluster -o jsonpath='{.status.platformStatus.aws.region}')

   Copy code to clipboard

3. Retrieve the security group IDs and store them in an array:

   $ AWS_SG_IDS=($(aws ec2 describe-instances --instance-ids ${INSTANCE_ID} \
     --query 'Reservations[*].Instances[*].SecurityGroups[*].GroupId' \
     --output text --region $AWS_REGION))

   Copy code to clipboard

4. For each security group ID, authorize the peer pods shim to access kata-agent communication, and set up the peer pods tunnel:

   $ for AWS_SG_ID in "${AWS_SG_IDS[@]}"; do \
     aws ec2 authorize-security-group-ingress --group-id $AWS_SG_ID --protocol tcp --port 15150 --source-group $AWS_SG_ID --region $AWS_REGION; \
     aws ec2 authorize-security-group-ingress --group-id $AWS_SG_ID --protocol tcp --port 9000 --source-group $AWS_SG_ID --region $AWS_REGION; \
   done

   Copy code to clipboard
   Note

   The aws ec2 authorize-security-group-ingress command is not idempotent. If the rules already exist, the command returns an InvalidPermission.Duplicate error and exits with a nonzero status. This causes the loop to stop before processing remaining security groups. If you are re-running this procedure, verify that the rules do not already exist before running the loop.

   The ports are now enabled.

Procedure

1. Retrieve and record the instance ID by running the following command:

   $ INSTANCE_ID=$(oc get nodes -l 'node-role.kubernetes.io/worker' \
     -o jsonpath='{.items[0].spec.providerID}' | sed 's#[^ ]*/##g')

   Copy code to clipboard

   Use this value to retrieve other values for the secret object.

2. Retrieve and record the AWS region by running the following command:

   $ AWS_REGION=$(oc get infrastructure/cluster \
     -o jsonpath='{.status.platformStatus.aws.region}') \
     && echo "AWS_REGION: \"$AWS_REGION\""

   Copy code to clipboard

3. Retrieve and record the AWS subnet ID by running the following command:

   $ AWS_SUBNET_ID=$(aws ec2 describe-instances --instance-ids ${INSTANCE_ID} \
     --query 'Reservations[*].Instances[*].SubnetId' --region ${AWS_REGION} \
       --output text) && echo "AWS_SUBNET_ID: \"$AWS_SUBNET_ID\""

   Copy code to clipboard

4. Retrieve and record the AWS VPC ID by running the following command:

   $ AWS_VPC_ID=$(aws ec2 describe-instances --instance-ids ${INSTANCE_ID} \
     --query 'Reservations[*].Instances[*].VpcId' --region ${AWS_REGION} \
       --output text) && echo "AWS_VPC_ID: \"$AWS_VPC_ID\""

   Copy code to clipboard

5. Retrieve and record the AWS security group IDs by running the following command:

   AWS_SG_IDS=$(aws ec2 describe-instances --instance-ids ${INSTANCE_ID} \
     --query 'Reservations[*].Instances[*].SecurityGroups[*].GroupId' \
     --region  $AWS_REGION --output json | jq -r '.[][][]' | paste -sd "," -) \
       && echo "AWS_SG_IDS: \"$AWS_SG_IDS\""

   Copy code to clipboard

Procedure

1. Create a peer-pods-cm.yaml manifest file according to the following example:

   apiVersion: v1
   kind: ConfigMap
   metadata:
     name: peer-pods-cm
     namespace: openshift-sandboxed-containers-operator
   data:
     CLOUD_PROVIDER: "aws"
     VXLAN_PORT: "9000"
     PROXY_TIMEOUT: "8m"
     PODVM_INSTANCE_TYPE: "t3.medium"
     PODVM_INSTANCE_TYPES: "t2.small,t2.medium,t3.large"
     PODVM_AMI_ID: "<podvm_ami_id>"
     AWS_REGION: "<aws_region>"
     AWS_SUBNET_ID: "<aws_subnet_id>"
     AWS_VPC_ID: "<aws_vpc_id>"
     AWS_SG_IDS: "<aws_sg_ids>"
     TAGS: "key1=value1,key2=value2"
     PEERPODS_LIMIT_PER_NODE: "10"
     ROOT_VOLUME_SIZE: "6"
     DISABLECVM: "true"

   Copy code to clipboard

   PODVM_INSTANCE_TYPE

   Defines the default instance type that is used if the instance type is not defined in the workload object.

   PODVM_INSTANCE_TYPES

   Specify the allowed instance types, without spaces, for creating the pod. You can define smaller instance types for workloads that need less memory and fewer CPUs or larger instance types for larger workloads.

   PODVM_AMI_ID

   This value is populated when you run the KataConfig custom resource (CR), using an AMI ID based on your cluster credentials. If you create your own AMI, specify the correct AMI ID.

   TAGS

   You can configure custom tags as key=value pairs for pod VM instances to track peer pod costs or to identify peer pods in different clusters.

   PEERPODS_LIMIT_PER_NODE

   Maximum number of peer pods per node. The default value is 10.

   ROOT_VOLUME_SIZE

   Root volume size in gigabytes for the pod VM. The default and minimum size is 6 GB.

2. Create the config map by running the following command:

   $ oc create -f peer-pods-cm.yaml

   Copy code to clipboard

Verification

1. Verify that the peer pods config map exists by running the following command:

   $ oc get configmap peer-pods-cm -n openshift-sandboxed-containers-operator

   Copy code to clipboard

   NAME           DATA   AGE
   peer-pods-cm   10     10s

   Copy code to clipboard

Procedure

1. Create an example-kataconfig.yaml manifest file according to the following example:

   apiVersion: kataconfiguration.openshift.io/v1
   kind: KataConfig
   metadata:
     name: example-kataconfig
   spec:
     enablePeerPods: true
     logLevel: info
   #  kataConfigPoolSelector:
   #    matchLabels:
   #      <label_key>: '<label_value>'

   Copy code to clipboard

   <label_key>: '<label_value>'

   Optional: If you have applied node labels to install kata-remote on specific nodes, specify the key and value, for example, kata-remote: 'true'.

2. Create the KataConfig CR by running the following command:

   $ oc create -f example-kataconfig.yaml

   Copy code to clipboard

   The new KataConfig CR is created and installs kata-remote as a runtime class on the worker nodes.

   Wait for the kata-remote installation to complete and the worker nodes to reboot before verifying the installation.

3. Optional: Monitor the installation progress by running the following command:

   $ watch "oc describe kataconfig | sed -n /^Status:/,/^Events/p"

   Copy code to clipboard

   When the status of all workers under kataNodes is installed and the condition InProgress is False without specifying a reason, the kata-remote is installed on the cluster.

Procedure

1. Obtain the config map that you created for the peer pods:

   $ oc get configmap peer-pods-cm -n openshift-sandboxed-containers-operator -o yaml

   Copy code to clipboard

2. Check the data section of the YAML output for the PODVM_AMI_ID parameter.

   If the PODVM_AMI_ID parameter is populated, the pod VM image was created successfully.

Troubleshooting

1. Retrieve the events log by running the following command:

   $ oc get events -n openshift-sandboxed-containers-operator --field-selector involvedObject.name=osc-podvm-image-creation

   Copy code to clipboard

2. Retrieve the job log by running the following command:

   $ oc logs -n openshift-sandboxed-containers-operator jobs/osc-podvm-image-creation

   Copy code to clipboard

Procedure

1. Create an initdata.toml file to modify the default policy:

   algorithm = "sha256"
   version = "0.1.0"
   
   [data]
   "aa.toml" = '''
   '''
   
   "policy.rego" = '''
   package agent_policy
   
   default AddARPNeighborsRequest := true
   default AddSwapRequest := true
   default CloseStdinRequest := true
   default CopyFileRequest := true
   default CreateContainerRequest := true
   default CreateSandboxRequest := true
   default DestroySandboxRequest := true
   default GetMetricsRequest := true
   default GetOOMEventRequest := true
   default GuestDetailsRequest := true
   default ListInterfacesRequest := true
   default ListRoutesRequest := true
   default MemHotplugByProbeRequest := true
   default OnlineCPUMemRequest := true
   default PauseContainerRequest := true
   default PullImageRequest := true
   default ReadStreamRequest := false
   default RemoveContainerRequest := true
   default RemoveStaleVirtiofsShareMountsRequest := true
   default ReseedRandomDevRequest := true
   default ResumeContainerRequest := true
   default SetGuestDateTimeRequest := true
   default SignalProcessRequest := true
   default StartContainerRequest := true
   default StartTracingRequest := true
   default StatsContainerRequest := true
   default StopTracingRequest := true
   default TtyWinResizeRequest := true
   default UpdateContainerRequest := true
   default UpdateEphemeralMountsRequest := true
   default UpdateInterfaceRequest := true
   default UpdateRoutesRequest := true
   default WaitProcessRequest := true
   default ExecProcessRequest := false
   default SetPolicyRequest := false
   default WriteStreamRequest := false
   '''

   Copy code to clipboard

   The default policy allows all API calls. Adjust the true or false values to customize the policy further based on your needs.

2. Convert the initdata.toml file to a Base64-encoded string by running the following command:

   $ cat initdata.toml | gzip | base64 -w0 > initdata.txt

   Copy code to clipboard

   Record this string to use in the pod manifest.

3. Add the Base64-encoded initdata string to the my-pod.yaml manifest:

   apiVersion: v1
   kind: Pod
   metadata:
     name: my-pod
     annotations:
       io.katacontainers.config.hypervisor.cc_init_data: <initdata_string>
   spec:
     runtimeClassName: kata-remote
     containers:
     - name: <container_name>
       image: registry.access.redhat.com/ubi9/ubi:latest
       command:
       - sleep
       - "36000"
       securityContext:
         privileged: false
         seccompProfile:
           type: RuntimeDefault

   Copy code to clipboard

4. Create the pod by running the following command:

   $ oc create -f my-pod.yaml

   Copy code to clipboard

Procedure

1. Optional: If the custom pod VM image is in a private registry, configure a pull secret:

   1. Set the NS variable to the namespace where you deploy your peer pods:

      $ NS=<namespace>

      Copy code to clipboard

   2. Copy the pull secret to the peer pod namespace:

      $ oc get secret pull-secret -n openshift-config -o yaml \
        | sed "s/namespace: openshift-config/namespace: ${NS}/" \
        | oc apply -n "${NS}" -f -

      Copy code to clipboard

      You can use the cluster pull secret, as in this example, or a custom pull secret.

   3. Optional: Link the pull secret to the default service account:

      $ oc secrets link default pull-secret --for=pull -n ${NS}

      Copy code to clipboard

2. Create a my-pod-manifest.yaml file according to the following example:

   apiVersion: v1
   kind: Pod
   metadata:
     name: my-pod-manifest
     annotations:
       io.katacontainers.config.hypervisor.image: "<custom_image_id>"
   spec:
     runtimeClassName: kata-remote
     containers:
     - name: <example_container>
       image: registry.access.redhat.com/ubi9/ubi:9.3
       command: ["sleep", "36000"]

   Copy code to clipboard

3. Create the pod by running the following command:

   $ oc create -f my-pod-manifest.yaml

   Copy code to clipboard

Verification

1. Verify that the pod is running with the custom peer pod VM image by running the following command:

   $ oc get pod my-pod-manifest -o jsonpath='{.metadata.annotations.io\.katacontainers\.config\.hypervisor\.image}'

   Copy code to clipboard
2. Confirm that the output matches the custom image ID you specified.

Procedure

1. Add spec.runtimeClassName: kata-remote to the manifest of each pod-templated workload object as in the following example:

   apiVersion: v1
   kind: <object>
   # ...
   spec:
     runtimeClassName: kata-remote
   # ...

   Copy code to clipboard

2. Optional: To override the default instance type, add the following annotation with an instance type that is defined in the peer pods config map:

   apiVersion: v1
   kind: <object>
   metadata:
     annotations:
       io.katacontainers.config.hypervisor.machine_type: <instance>
   # ...

   Copy code to clipboard

3. Optional: To use an automatic instance type, add the following annotations:

   apiVersion: v1
   kind: <Pod>
   metadata:
     annotations:
       io.katacontainers.config.hypervisor.default_vcpus: <vcpus>
       io.katacontainers.config.hypervisor.default_memory: <memory>
   # ...

   Copy code to clipboard

   The workload runs on an automatic instance type based on the amount of memory available.

4. Apply the changes to the workload object by running the following command:

   $ oc apply -f <object.yaml>

   Copy code to clipboard

   OpenShift Container Platform creates the workload object and begins scheduling it.

Procedure

1. Create an osc-subscription.yaml manifest file:

   apiVersion: operators.coreos.com/v1alpha1
   kind: Subscription
   metadata:
     name: sandboxed-containers-operator
     namespace: openshift-sandboxed-containers-operator
   spec:
     channel: stable
     installPlanApproval: Automatic
     name: sandboxed-containers-operator
     source: redhat-operators
     sourceNamespace: openshift-marketplace
     startingCSV: sandboxed-containers-operator.v1.13.1

   Copy code to clipboard

2. Create the subscription by running the following command:

   $ oc create -f osc-subscription.yaml

   Copy code to clipboard

Verification

1. Verify that the Operator upgrade is complete by running the following command:

   $ oc get csv -n openshift-sandboxed-containers-operator

   Copy code to clipboard

   This command can take several minutes to complete.

2. Watch the upgrade progress by running the following command:

   $ watch oc get csv -n openshift-sandboxed-containers-operator

   Copy code to clipboard

   NAME                             DISPLAY                                  VERSION   REPLACES    PHASE
   openshift-sandboxed-containers   openshift-sandboxed-containers-operator  1.13.1    1.13.0      Succeeded

   Copy code to clipboard

   The upgrade is complete when the PHASE column shows Succeeded for the new version.

Procedure

1. Delete the KataConfig CR by running the following command:

   $ oc delete kataconfig example-kataconfig

   Copy code to clipboard

   The OpenShift sandboxed containers Operator removes all resources that were initially created to enable the runtime on your cluster.

   Important

   When you delete the KataConfig CR, the command-line interface (CLI) stops responding until all worker nodes reboot. You must wait for the deletion process to complete before performing the verification.

Verification

1. Confirm that the KataConfig CR no longer exists by running the following command:

   $ oc get kataconfig example-kataconfig

   Copy code to clipboard

   Error from server (NotFound): kataconfigs.kataconfiguration.openshift.io "example-kataconfig" not found

   Copy code to clipboard

Procedure

1. Obtain the value of the PODVM_AMI_ID in the peer pods config map by running the following command:

   $ oc get configmap -n openshift-sandboxed-containers-operator peer-pods-cm -o jsonpath="{.data.PODVM_AMI_ID}"

   Copy code to clipboard

2. If the value is not empty, update the value and patch the config map by running the following command:

   $ oc patch configmap peer-pods-cm -n openshift-sandboxed-containers-operator -p '{"data":{"PODVM_AMI_ID":""}}'

   Copy code to clipboard

Procedure

1. Create an example-kataconfig.yaml manifest file according to the following example:

   apiVersion: kataconfiguration.openshift.io/v1
   kind: KataConfig
   metadata:
     name: example-kataconfig
   spec:
     enablePeerPods: true
     logLevel: info
   #  kataConfigPoolSelector:
   #    matchLabels:
   #      <label_key>: '<label_value>'

   Copy code to clipboard

   <label_key>: '<label_value>'

   Optional: If you have applied node labels to install kata-remote on specific nodes, specify the key and value, for example, kata-remote: 'true'.

2. Create the KataConfig CR by running the following command:

   $ oc create -f example-kataconfig.yaml

   Copy code to clipboard

   The new KataConfig CR is created and installs kata-remote as a runtime class on the worker nodes.

   Wait for the kata-remote installation to complete and the worker nodes to reboot before verifying the installation.

3. Optional: Monitor the installation progress by running the following command:

   $ watch "oc describe kataconfig | sed -n /^Status:/,/^Events/p"

   Copy code to clipboard

   When the status of all workers under kataNodes is installed and the condition InProgress is False without specifying a reason, the kata-remote is installed on the cluster.

Procedure

1. Search for the pods by running the following command:

   $ oc get pods -A -o json | jq -r '.items[] |
     select(.spec.runtimeClassName == "kata-remote").metadata.name'

   Copy code to clipboard

2. Delete each pod by running the following command:

   $ oc delete pod <pod>

   Copy code to clipboard

Verification

1. Verify that the pods using the kata-remote runtime class are no longer running by running the following command:

   $ oc get pods -A -o json | jq -r '.items[] | select(.spec.runtimeClassName == "kata-remote").metadata.name'

   Copy code to clipboard

   Confirm that the command returns no output.

Procedure

1. Delete the KataConfig CR by running the following command:

   $ oc delete kataconfig example-kataconfig

   Copy code to clipboard

   The OpenShift sandboxed containers Operator removes all resources that were initially created to enable the runtime on your cluster.

   Important

   When you delete the KataConfig CR, the command-line interface (CLI) stops responding until all worker nodes reboot. You must wait for the deletion process to complete before performing the verification.

Verification

1. Confirm that the KataConfig CR no longer exists by running the following command:

   $ oc get kataconfig example-kataconfig

   Copy code to clipboard

   Error from server (NotFound): kataconfigs.kataconfiguration.openshift.io "example-kataconfig" not found

   Copy code to clipboard

Procedure

1. Delete the subscription by running the following command:

   $ oc delete subscription sandboxed-containers-operator -n openshift-sandboxed-containers-operator

   Copy code to clipboard

2. Delete the namespace by running the following command:

   $ oc delete namespace openshift-sandboxed-containers-operator

   Copy code to clipboard

Procedure

1. Delete the KataConfig CRD by running the following command:

   $ oc delete crd kataconfigs.kataconfiguration.openshift.io

   Copy code to clipboard

Verification

1. Confirm that the KataConfig CRD no longer exists by running the following command:

   $ oc get crd kataconfigs.kataconfiguration.openshift.io

   Copy code to clipboard

   Error from server (NotFound): customresourcedefinitions.apiextensions.k8s.io "kataconfigs.kataconfiguration.openshift.io" not found

   Copy code to clipboard

Procedure

1. In the OpenShift Container Platform web console, navigate to Observe → Metrics.

2. In the input field, enter the query for the metric you want to observe.

   All kata-related metrics begin with kata. Typing kata displays a list of all available kata metrics.

   The metrics from your query are visualized on the page.

Procedure

1. Change the logLevel field in your existing KataConfig CR to debug:

   $ oc patch kataconfig <kataconfig> --type merge --patch '{"spec":{"logLevel":"debug"}}'

   Copy code to clipboard

2. Monitor the kata-oc machine config pool until the value of UPDATED is True, indicating that all worker nodes are updated:

   $ oc get mcp kata-oc

   Copy code to clipboard

   Example output

   NAME     CONFIG                 UPDATED  UPDATING  DEGRADED  MACHINECOUNT  READYMACHINECOUNT  UPDATEDMACHINECOUNT  DEGRADEDMACHINECOUNT  AGE
   kata-oc  rendered-kata-oc-169   False    True      False     3             1                  1                    0                     9h

   Copy code to clipboard

Verification

1. Start a debug session with a node in the machine config pool:

   $ oc debug node/<node_name>

   Copy code to clipboard

2. Change the root directory to /host:

   # chroot /host

   Copy code to clipboard

3. Verify the changes in the crio.conf file:

   # crio config | egrep 'log_level

   Copy code to clipboard

   Example output

   log_level = "debug"

   Copy code to clipboard