Soon, all Red Hat OpenShift documentation will only be available on docs.redhat.com, the official location of all Red Hat product documentation. Get familiar with the updated experience today!
  1. Documentation
×

Introduction to OpenShift updates

With OKD 4, you can update an OKD cluster with a single operation by using the web console or the OpenShift CLI (oc). Platform administrators can view new update options either by going to Administration → Cluster Settings in the web console or by looking at the output of the oc adm upgrade command.

Red Hat hosts a public OpenShift Update Service (OSUS), which serves a graph of update possibilities based on the OKD release images in the official registry. The graph contains update information for any public OCP release. OKD clusters are configured to connect to the OSUS by default, and the OSUS responds to clusters with information about known update targets.

An update begins when either a cluster administrator or an automatic update controller edits the custom resource (CR) of the Cluster Version Operator (CVO) with a new version. To reconcile the cluster with the newly specified version, the CVO retrieves the target release image from an image registry and begins to apply changes to the cluster.

Operators previously installed through Operator Lifecycle Manager (OLM) follow a different process for updates. See Updating installed Operators for more information.

The target release image contains manifest files for all cluster components that form a specific OCP version. When updating the cluster to a new version, the CVO applies manifests in separate stages called Runlevels. Most, but not all, manifests support one of the cluster Operators. As the CVO applies a manifest to a cluster Operator, the Operator might perform update tasks to reconcile itself with its new specified version.

The CVO monitors the state of each applied resource and the states reported by all cluster Operators. The CVO only proceeds with the update when all manifests and cluster Operators in the active Runlevel reach a stable condition. After the CVO updates the entire control plane through this process, the Machine Config Operator (MCO) updates the operating system and configuration of every node in the cluster.

About the OpenShift Update Service

The OpenShift Update Service (OSUS) provides update recommendations to OKD, including Fedora CoreOS (FCOS). It provides a graph, or diagram, that contains the vertices of component Operators and the edges that connect them. The edges in the graph show which versions you can safely update to. The vertices are update payloads that specify the intended state of the managed cluster components.

The Cluster Version Operator (CVO) in your cluster checks with the OpenShift Update Service to see the valid updates and update paths based on current component versions and information in the graph. When you request an update, the CVO uses the corresponding release image to update your cluster. The release artifacts are hosted in Quay as container images.

To allow the OpenShift Update Service to provide only compatible updates, a release verification pipeline drives automation. Each release artifact is verified for compatibility with supported cloud platforms and system architectures, as well as other component packages. After the pipeline confirms the suitability of a release, the OpenShift Update Service notifies you that it is available.

The OpenShift Update Service displays all recommended updates for your current cluster. If an update path is not recommended by the OpenShift Update Service, it might be because of a known issue related to the update path, such as incompatibility or availability.

Two controllers run during continuous update mode. The first controller continuously updates the payload manifests, applies the manifests to the cluster, and outputs the controlled rollout status of the Operators to indicate whether they are available, upgrading, or failed. The second controller polls the OpenShift Update Service to determine if updates are available.

Only updating to a newer version is supported. Reverting or rolling back your cluster to a previous version is not supported. If your update fails, contact Red Hat support.

During the update process, the Machine Config Operator (MCO) applies the new configuration to your cluster machines. The MCO cordons the number of nodes specified by the maxUnavailable field on the machine configuration pool and marks them unavailable. By default, this value is set to 1. The MCO updates the affected nodes alphabetically by zone, based on the topology.kubernetes.io/zone label. If a zone has more than one node, the oldest nodes are updated first. For nodes that do not use zones, such as in bare metal deployments, the nodes are updated by age, with the oldest nodes updated first. The MCO updates the number of nodes as specified by the maxUnavailable field on the machine configuration pool at a time. The MCO then applies the new configuration and reboots the machine.

The default setting for maxUnavailable is 1 for all the machine config pools in OKD. It is recommended to not change this value and update one control plane node at a time. Do not change this value to 3 for the control plane pool.

If you use Fedora machines as workers, the MCO does not update the kubelet because you must update the OpenShift API on the machines first.

With the specification for the new version applied to the old kubelet, the Fedora machine cannot return to the Ready state. You cannot complete the update until the machines are available. However, the maximum number of unavailable nodes is set to ensure that normal cluster operations can continue with that number of machines out of service.

The OpenShift Update Service is composed of an Operator and one or more application instances.

Understanding cluster Operator condition types

The status of cluster Operators includes their condition type, which informs you of the current state of your Operator’s health. The following definitions cover a list of some common ClusterOperator condition types. Operators that have additional condition types and use Operator-specific language have been omitted.

The Cluster Version Operator (CVO) is responsible for collecting the status conditions from cluster Operators so that cluster administrators can better understand the state of the OKD cluster.

  • Available: The condition type Available indicates that an Operator is functional and available in the cluster. If the status is False, at least one part of the operand is non-functional and the condition requires an administrator to intervene.

  • Progressing: The condition type Progressing indicates that an Operator is actively rolling out new code, propagating configuration changes, or otherwise moving from one steady state to another.

    Operators do not report the condition type Progressing as True when they are reconciling a previous known state. If the observed cluster state has changed and the Operator is reacting to it, then the status reports back as True, since it is moving from one steady state to another.

  • Degraded: The condition type Degraded indicates that an Operator has a current state that does not match its required state over a period of time. The period of time can vary by component, but a Degraded status represents persistent observation of an Operator’s condition. As a result, an Operator does not fluctuate in and out of the Degraded state.

    There might be a different condition type if the transition from one state to another does not persist over a long enough period to report Degraded. An Operator does not report Degraded during the course of a normal update. An Operator may report Degraded in response to a persistent infrastructure failure that requires eventual administrator intervention.

    This condition type is only an indication that something may need investigation and adjustment. As long as the Operator is available, the Degraded condition does not cause user workload failure or application downtime.

  • Upgradeable: The condition type Upgradeable indicates whether the Operator is safe to update based on the current cluster state. The message field contains a human-readable description of what the administrator needs to do for the cluster to successfully update. The CVO allows updates when this condition is True, Unknown or missing.

    When the Upgradeable status is False, only minor updates are impacted, and the CVO prevents the cluster from performing impacted updates unless forced.

Understanding cluster version condition types

The Cluster Version Operator (CVO) monitors cluster Operators and other components, and is responsible for collecting the status of both the cluster version and its Operators. This status includes the condition type, which informs you of the health and current state of the OKD cluster.

In addition to Available, Progressing, and Upgradeable, there are condition types that affect cluster versions and Operators.

  • Failing: The cluster version condition type Failing indicates that a cluster cannot reach its desired state, is unhealthy, and requires an administrator to intervene.

  • Invalid: The cluster version condition type Invalid indicates that the cluster version has an error that prevents the server from taking action. The CVO only reconciles the current state as long as this condition is set.

  • RetrievedUpdates: The cluster version condition type RetrievedUpdates indicates whether or not available updates have been retrieved from the upstream update server. The condition is Unknown before retrieval, False if the updates either recently failed or could not be retrieved, or True if the availableUpdates field is both recent and accurate.

  • ReleaseAccepted: The cluster version condition type ReleaseAccepted with a True status indicates that the requested release payload was successfully loaded without failure during image verification and precondition checking.

  • ImplicitlyEnabledCapabilities: The cluster version condition type ImplicitlyEnabledCapabilities with a True status indicates that there are enabled capabilities that the user is not currently requesting through spec.capabilities. The CVO does not support disabling capabilities if any associated resources were previously managed by the CVO.

Common terms

Control plane

The control plane, which is composed of control plane machines, manages the OKD cluster. The control plane machines manage workloads on the compute machines, which are also known as worker machines.

Cluster Version Operator

The Cluster Version Operator (CVO) starts the update process for the cluster. It checks with OSUS based on the current cluster version and retrieves the graph which contains available or possible update paths.

Machine Config Operator

The Machine Config Operator (MCO) is a cluster-level Operator that manages the operating system and machine configurations. Through the MCO, platform administrators can configure and update systemd, CRI-O and Kubelet, the kernel, NetworkManager, and other system features on the worker nodes.

OpenShift Update Service

The OpenShift Update Service (OSUS) provides over-the-air updates to OKD, including to Fedora CoreOS (FCOS). It provides a graph, or diagram, that contains the vertices of component Operators and the edges that connect them.

Channels

Channels declare an update strategy tied to minor versions of OKD. The OSUS uses this configured strategy to recommend update edges consistent with that strategy.

Recommended update edge

A recommended update edge is a recommended update between OKD releases. Whether a given update is recommended can depend on the cluster’s configured channel, current version, known bugs, and other information. OSUS communicates the recommended edges to the CVO, which runs in every cluster.

Additional resources