This is a cache of https://docs.okd.io/4.13/machine_management/modifying-machineset.html. It is a snapshot of the page at 2024-11-24T00:34:17.169+0000.
Modifying a compute machine set | Machine management | OKD 4.13
  1. Documentation
×

Modifying a compute machine set

You can modify a compute machine set, such as adding labels, changing the instance type, or changing block storage.

On oVirt, you can also change a compute machine set to provision new nodes on a different storage domain.

If you need to scale a compute machine set without making other changes, see Manually scaling a compute machine set.

Modifying a compute machine set by using the CLI

You can modify the configuration of a compute machine set, and then propagate the changes to the machines in your cluster by using the CLI.

By updating the compute machine set configuration, you can enable features or change the properties of the machines it creates. When you modify a compute machine set, your changes only apply to compute machines that are created after you save the updated MachineSet custom resource (CR). The changes do not affect existing machines.

Changes made in the underlying cloud provider are not reflected in the Machine or MachineSet CRs. To adjust instance configuration in cluster-managed infrastructure, use the cluster-side resources.

You can replace the existing machines with new ones that reflect the updated configuration by scaling the compute machine set to create twice the number of replicas and then scaling it down to the original number of replicas.

If you need to scale a compute machine set without making other changes, you do not need to delete the machines.

By default, the OKD router pods are deployed on compute machines. Because the router is required to access some cluster resources, including the web console, do not scale the compute machine set to 0 unless you first relocate the router pods.
Prerequisites

  • Your OKD cluster uses the Machine API.

  • You are logged in to the cluster as an administrator by using the OpenShift CLI (oc).

Procedure

  1. List the compute machine sets in your cluster by running the following command:

    $ oc get machinesets.machine.openshift.io -n openshift-machine-api
    Example output
    NAME                           DESIRED   CURRENT   READY   AVAILABLE   AGE
<compute_machine_set_name_1>   1         1         1       1           55m
<compute_machine_set_name_2>   1         1         1       1           55m

  2. Edit a compute machine set by running the following command:

    $ oc edit machinesets.machine.openshift.io <machine_set_name> \
  -n openshift-machine-api

  3. Note the value of the spec.replicas field, as you need it when scaling the machine set to apply the changes.

    apiVersion: machine.openshift.io/v1beta1
kind: MachineSet
metadata:
  name: <machine_set_name>
  namespace: openshift-machine-api
spec:
  replicas: 2 (1)
# ...
    1 The examples in this procedure show a compute machine set that has a replicas value of 2.

  4. Update the compute machine set CR with the configuration options that you want and save your changes.

  5. List the machines that are managed by the updated compute machine set by running the following command:

    $ oc get machines.machine.openshift.io \
  -n openshift-machine-api \
  -l machine.openshift.io/cluster-api-machineset=<machine_set_name>
    Example output for an AWS cluster
    NAME                        PHASE     TYPE         REGION      ZONE         AGE
<machine_name_original_1>   Running   m6i.xlarge   us-west-1   us-west-1a   4h
<machine_name_original_2>   Running   m6i.xlarge   us-west-1   us-west-1a   4h

  6. For each machine that is managed by the updated compute machine set, set the delete annotation by running the following command:

    $ oc annotate machine.machine.openshift.io/<machine_name_original_1> \
  -n openshift-machine-api \
  machine.openshift.io/delete-machine="true"

  7. Scale the compute machine set to twice the number of replicas by running the following command:

    $ oc scale --replicas=4 \(1)
  machineset.machine.openshift.io <machine_set_name> \
  -n openshift-machine-api
    1 The original example value of 2 is doubled to 4.

  8. List the machines that are managed by the updated compute machine set by running the following command:

    $ oc get machines.machine.openshift.io \
  -n openshift-machine-api \
  -l machine.openshift.io/cluster-api-machineset=<machine_set_name>
    Example output for an AWS cluster
    NAME                        PHASE          TYPE         REGION      ZONE         AGE
<machine_name_original_1>   Running        m6i.xlarge   us-west-1   us-west-1a   4h
<machine_name_original_2>   Running        m6i.xlarge   us-west-1   us-west-1a   4h
<machine_name_updated_1>    Provisioned    m6i.xlarge   us-west-1   us-west-1a   55s
<machine_name_updated_2>    Provisioning   m6i.xlarge   us-west-1   us-west-1a   55s

    When the new machines are in the Running phase, you can scale the compute machine set to the original number of replicas.

  9. Scale the compute machine set to the original number of replicas by running the following command:

    $ oc scale --replicas=2 \(1)
  machineset.machine.openshift.io <machine_set_name> \
  -n openshift-machine-api
    1 The original example value of 2.
Verification

  • To verify that a machine created by the updated machine set has the correct configuration, examine the relevant fields in the CR for one of the new machines by running the following command:

    $ oc describe machine.machine.openshift.io <machine_name_updated_1> \
  -n openshift-machine-api

  • To verify that the compute machines without the updated configuration are deleted, list the machines that are managed by the updated compute machine set by running the following command:

    $ oc get machines.machine.openshift.io \
  -n openshift-machine-api \
  -l machine.openshift.io/cluster-api-machineset=<machine_set_name>
    Example output while deletion is in progress for an AWS cluster
    NAME                        PHASE           TYPE         REGION      ZONE         AGE
<machine_name_original_1>   Deleting        m6i.xlarge   us-west-1   us-west-1a   4h
<machine_name_original_2>   Deleting        m6i.xlarge   us-west-1   us-west-1a   4h
<machine_name_updated_1>    Running         m6i.xlarge   us-west-1   us-west-1a   5m41s
<machine_name_updated_2>    Running         m6i.xlarge   us-west-1   us-west-1a   5m41s
    Example output when deletion is complete for an AWS cluster
    NAME                        PHASE           TYPE         REGION      ZONE         AGE
<machine_name_updated_1>    Running         m6i.xlarge   us-west-1   us-west-1a   6m30s
<machine_name_updated_2>    Running         m6i.xlarge   us-west-1   us-west-1a   6m30s
Additional resources

Migrating nodes to a different storage domain on oVirt

You can migrate the OKD control plane and compute nodes to a different storage domain in a oVirt cluster.

Migrating compute nodes to a different storage domain in oVirt

Prerequisites

  • You are logged in to the Engine.

  • You have the name of the target storage domain.

Procedure

  1. Identify the virtual machine template by running the following command:

    $ oc get -o jsonpath='{.items[0].spec.template.spec.providerSpec.value.template_name}{"\n"}' machineset -A

  2. Create a new virtual machine in the Engine, based on the template you identified. Leave all other settings unchanged. For details, see Creating a Virtual Machine Based on a Template in the Red Hat Virtualization Virtual Machine Management Guide.

    You do not need to start the new virtual machine.

  3. Create a new template from the new virtual machine. Specify the target storage domain under Target. For details, see Creating a Template in the Red Hat Virtualization Virtual Machine Management Guide.

  4. Add a new compute machine set to the OKD cluster with the new template.

    1. Get the details of the current compute machine set by running the following command:

      $ oc get machineset -o yaml

    2. Use these details to create a compute machine set. For more information see Creating a compute machine set.

      Enter the new virtual machine template name in the template_name field. Use the same template name you used in the New template dialog in the Engine.

    3. Note the names of both the old and new compute machine sets. You need to refer to them in subsequent steps.

  5. Migrate the workloads.

    1. Scale up the new compute machine set. For details on manually scaling compute machine sets, see Scaling a compute machine set manually.

      OKD moves the pods to an available worker when the old machine is removed.

    2. Scale down the old compute machine set.

  6. Remove the old compute machine set by running the following command:

    $ oc delete machineset <machineset-name>
Additional resources

Migrating control plane nodes to a different storage domain on oVirt

OKD does not manage control plane nodes, so they are easier to migrate than compute nodes. You can migrate them like any other virtual machine on oVirt.

Perform this procedure for each node separately.

Prerequisites

  • You are logged in to the Engine.

  • You have identified the control plane nodes. They are labeled master in the Engine.

Procedure

  1. Select the virtual machine labeled master.

  2. Shut down the virtual machine.

  3. Click the Disks tab.

  4. Click the virtual machine’s disk.

  5. Click More Actionskebab and select Move.

  6. Select the target storage domain and wait for the migration process to complete.

  7. Start the virtual machine.

  8. Verify that the OKD cluster is stable:

    $ oc get nodes

    The output should display the node with the status Ready.

  9. Repeat this procedure for each control plane node.