×

Due to fundamental Kubernetes design, all OpenShift Container Platform updates between minor versions must be serialized. You must update from OpenShift Container Platform <4.y> to <4.y+1>, and then to <4.y+2>. You cannot update from OpenShift Container Platform <4.y> to <4.y+2> directly. However, administrators who want to update between two even-numbered minor versions can do so incurring only a single reboot of non-control plane hosts.

This update was previously known as an EUS-to-EUS update and is now referred to as a Control Plane Only update. These updates are only viable between even-numbered minor versions of OpenShift Container Platform.

There are several caveats to consider when attempting a Control Plane Only update.

  • Control Plane Only updates are only offered after updates between all versions involved have been made available in stable channels.

  • If you encounter issues during or after updating to the odd-numbered minor version but before updating to the next even-numbered version, then remediation of those issues may require that non-control plane hosts complete the update to the odd-numbered version before moving forward.

  • You can do a partial update by updating the worker or custom pool nodes to accommodate the time it takes for maintenance.

  • You can complete the update process during multiple maintenance windows by pausing at intermediate steps. However, plan to complete the entire update within 60 days. This is critical to ensure that normal cluster automation processes are completed.

  • Until the machine config pools are unpaused and the update is complete, some features and bugs fixes in <4.y+1> and <4.y+2> of OpenShift Container Platform are not available.

  • All the clusters might update using EUS channels for a conventional update without pools paused, but only clusters with non control-plane MachineConfigPools objects can do Control Plane Only updates with pools paused.

Performing a Control Plane Only update

The following procedure pauses all non-master machine config pools and performs updates from OpenShift Container Platform <4.y> to <4.y+1> to <4.y+2>, then unpauses the previously paused machine config pools. Following this procedure reduces the total update duration and the number of times worker nodes are restarted.

Prerequisites
  • Review the release notes for OpenShift Container Platform <4.y+1> and <4.y+2>.

  • Review the release notes and product lifecycles for any layered products and Operator Lifecycle Manager (OLM) Operators. Some may require updates either before or during a Control Plane Only update.

  • Ensure that you are familiar with version-specific prerequisites, such as the removal of deprecated APIs, that are required prior to updating from OpenShift Container Platform <4.y+1> to <4.y+2>.

  • If your cluster uses in-tree vSphere volumes, update vSphere to version 7.0u3L+ or 8.0u2+.

    If you do not update vSphere to 7.0u3L+ or 8.0u2+ before initiating an OpenShift Container Platform update, known issues might occur with your cluster after the update. For more information, see Known Issues with OpenShift 4.12 to 4.13 or 4.13 to 4.14 vSphere CSI Storage Migration.

Performing a Control Plane Only update using the web console

Prerequisites
  • Verify that machine config pools are unpaused.

  • Have access to the web console as a user with admin privileges.

Procedure
  1. Using the Administrator perspective on the web console, update any Operator Lifecycle Manager (OLM) Operators to the versions that are compatible with your intended updated version. You can find more information on how to perform this action in "Updating installed Operators"; see "Additional resources".

  2. Verify that all machine config pools display a status of Up to date and that no machine config pool displays a status of UPDATING.

    To view the status of all machine config pools, click ComputeMachineConfigPools and review the contents of the Update status column.

    If your machine config pools have an Updating status, please wait for this status to change to Up to date. This process could take several minutes.

  3. Set your channel to eus-<4.y+2>.

    To set your channel, click AdministrationCluster SettingsChannel. You can edit your channel by clicking on the current hyperlinked channel.

  4. Pause all worker machine pools except for the master pool. You can perform this action on the MachineConfigPools tab under the Compute page. Select the vertical ellipses next to the machine config pool you’d like to pause and click Pause updates.

  5. Update to version <4.y+1> and complete up to the Save step. You can find more information on how to perform these actions in "Updating a cluster by using the web console"; see "Additional resources".

  6. Ensure that the <4.y+1> updates are complete by viewing the Last completed version of your cluster. You can find this information on the Cluster Settings page under the Details tab.

  7. If necessary, update your OLM Operators by using the Administrator perspective on the web console. You can find more information on how to perform these actions in "Updating installed Operators"; see "Additional resources".

  8. Update to version <4.y+2> and complete up to the Save step. You can find more information on how to perform these actions in "Updating a cluster by using the web console"; see "Additional resources".

  9. Ensure that the <4.y+2> update is complete by viewing the Last completed version of your cluster. You can find this information on the Cluster Settings page under the Details tab.

  10. Unpause all previously paused machine config pools. You can perform this action on the MachineConfigPools tab under the Compute page. Select the vertical ellipses next to the machine config pool you’d like to unpause and click Unpause updates.

    If pools are paused, the cluster is not permitted to upgrade to any future minor versions, and some maintenance tasks are inhibited. This puts the cluster at risk for future degradation.

  11. Verify that your previously paused pools are updated and that your cluster has completed the update to version <4.y+2>.

    You can verify that your pools have updated on the MachineConfigPools tab under the Compute page by confirming that the Update status has a value of Up to date.

    When you update a cluster that contains Red Hat Enterprise Linux (RHEL) compute machines, those machines temporarily become unavailable during the update process. You must run the upgrade playbook against each RHEL machine as it enters the NotReady state for the cluster to finish updating. For more information, see "Updating a cluster that includes RHEL compute machines" in the additional resources section.

    You can verify that your cluster has completed the update by viewing the Last completed version of your cluster. You can find this information on the Cluster Settings page under the Details tab.

Performing a Control Plane Only update using the CLI

Prerequisites
  • Verify that machine config pools are unpaused.

  • Update the OpenShift CLI (oc) to the target version before each update.

It is highly discouraged to skip this prerequisite. If the OpenShift CLI (oc) is not updated to the target version before your update, unexpected issues may occur.

Procedure
  1. Using the Administrator perspective on the web console, update any Operator Lifecycle Manager (OLM) Operators to the versions that are compatible with your intended updated version. You can find more information on how to perform this action in "Updating installed Operators"; see "Additional resources".

  2. Verify that all machine config pools display a status of UPDATED and that no machine config pool displays a status of UPDATING. To view the status of all machine config pools, run the following command:

    $ oc get mcp
    Example output
    NAME     CONFIG                                         	UPDATED   UPDATING
    master   rendered-master-ecbb9582781c1091e1c9f19d50cf836c       True  	  False
    worker   rendered-worker-00a3f0c68ae94e747193156b491553d5       True  	  False
  3. Your current version is <4.y>, and your intended version to update is <4.y+2>. Change to the eus-<4.y+2> channel by running the following command:

    $ oc adm upgrade channel eus-<4.y+2>

    If you receive an error message indicating that eus-<4.y+2> is not one of the available channels, this indicates that Red Hat is still rolling out EUS version updates. This rollout process generally takes 45-90 days starting at the GA date.

  4. Pause all worker machine pools except for the master pool by running the following command:

    $ oc patch mcp/worker --type merge --patch '{"spec":{"paused":true}}'

    You cannot pause the master pool.

  5. Update to the latest version by running the following command:

    $ oc adm upgrade --to-latest
    Example output
    Updating to latest version <4.y+1.z>
  6. Review the cluster version to ensure that the updates are complete by running the following command:

    $ oc adm upgrade
    Example output
    Cluster version is <4.y+1.z>
    ...
  7. Update to version <4.y+2> by running the following command:

    $ oc adm upgrade --to-latest
  8. Retrieve the cluster version to ensure that the <4.y+2> updates are complete by running the following command:

    $ oc adm upgrade
    Example output
    Cluster version is <4.y+2.z>
    ...
  9. To update your worker nodes to <4.y+2>, unpause all previously paused machine config pools by running the following command:

    $ oc patch mcp/worker --type merge --patch '{"spec":{"paused":false}}'

    If pools are not unpaused, the cluster is not permitted to update to any future minor versions, and some maintenance tasks are inhibited. This puts the cluster at risk for future degradation.

  10. Verify that your previously paused pools are updated and that the update to version <4.y+2> is complete by running the following command:

    $ oc get mcp

    When you update a cluster that contains Red Hat Enterprise Linux (RHEL) compute machines, those machines temporarily become unavailable during the update process. You must run the upgrade playbook against each RHEL machine as it enters the NotReady state for the cluster to finish updating. For more information, see "Updating a cluster that includes RHEL compute machines" in the additional resources section.

    Example output
    NAME 	   CONFIG                                            UPDATED     UPDATING
    master   rendered-master-52da4d2760807cb2b96a3402179a9a4c    True  	 False
    worker   rendered-worker-4756f60eccae96fb9dcb4c392c69d497    True 	 False

Performing a Control Plane Only update for layered products and Operators installed through Operator Lifecycle Manager

In addition to the Control Plane Only update steps mentioned for the web console and CLI, there are additional steps to consider when performing Control Plane Only updates for clusters with the following:

  • Layered products

  • Operators installed through Operator Lifecycle Manager (OLM)

What is a layered product?

Layered products refer to products that are made of multiple underlying products that are intended to be used together and cannot be broken into individual subscriptions. For examples of layered OpenShift Container Platform products, see Layered Offering On OpenShift.

As you perform a Control Plane Only update for the clusters of layered products and those of Operators that have been installed through OLM, you must complete the following:

  1. You have updated all Operators previously installed through Operator Lifecycle Manager (OLM) to a version that is compatible with your target release. Updating the Operators ensures they have a valid update path when the default OperatorHub catalogs switch from the current minor version to the next during a cluster update. See "Updating installed Operators" in the "Additional resources" section for more information on how to check compatibility and, if necessary, update the installed Operators.

  2. Confirm the cluster version compatibility between the current and intended Operator versions. You can verify which versions your OLM Operators are compatible with by using the Red Hat OpenShift Container Platform Operator Update Information Checker.

As an example, here are the steps to perform a Control Plane Only update from <4.y> to <4.y+2> for OpenShift Data Foundation (ODF). This can be done through the CLI or web console. For information on how to update clusters through your desired interface, see Performing a Control Plane Only update using the web console and Performing a Control Plane Only update using the CLI in "Additional resources".

Example workflow
  1. Pause the worker machine pools.

  2. Update OpenShift <4.y> → OpenShift <4.y+1>.

  3. Update ODF <4.y> → ODF <4.y+1>.

  4. Update OpenShift <4.y+1> → OpenShift <4.y+2>.

  5. Update to ODF <4.y+2>.

  6. Unpause the worker machine pools.

The update to ODF <4.y+2> can happen before or after worker machine pools have been unpaused.