Skip to content

Rook Upgrades

This guide will walk through the steps to upgrade the software in a Rook cluster from one version to the next. This guide focuses on updating the Rook version for the management layer, while the Ceph upgrade guide focuses on updating the data layer.

Upgrades for both the operator and for Ceph are entirely automated except where Rook's permissions need to be explicitly updated by an admin or when incompatibilities need to be addressed manually due to customizations.

We welcome feedback and opening issues!

Supported Versions

This guide is for upgrading from Rook v1.20.x to Rook v1.21.x.

Please refer to the upgrade guides from previous releases for supported upgrade paths. Rook upgrades are only supported between official releases.

For a guide to upgrade previous versions of Rook, please refer to the version of documentation for those releases.

Important

Rook releases from master are expressly unsupported. It is strongly recommended to use official releases of Rook. Unreleased versions from the master branch are subject to changes and incompatibilities that will not be supported in the official releases. Builds from the master branch can have functionality changed or removed at any time without compatibility support and without prior notice.

Breaking changes in v1.21

  • Helm OCI chart tags no longer include the v prefix (e.g., 1.21.0 instead of v1.21.0). Update any scripts or tooling that reference the chart by tag.
  • Ceph msgrv2 is required by default.
    • Msgrv2 requires the 5.11 kernel. If you have an older kernel, disable the msgrv2 protocol with the CephCluster CR setting network.connections.requireMsgr2: false. If using the helm chart, this same value is applied under the cephClusterSpec of the values.
    • To preserve compatibility, existing volume mounts will continue to use the msgrv1 protocol until they are drained and remounted.
    • Ceph mons will not exclusively require msgrv2 until new mons are deployed during mon failover.
  • The minimum supported Kubernetes version is v1.32.

Considerations

With this upgrade guide, there are a few notes to consider:

  • WARNING: Upgrading a Rook cluster is not without risk. There may be unexpected issues or obstacles that damage the integrity and health of the storage cluster, including data loss.
  • The Rook cluster's storage may be unavailable for short periods during the upgrade process for both Rook operator updates and for Ceph version updates.
  • Read this document in full before undertaking the cluster upgrade.

Patch Release Upgrades

Unless otherwise noted due to extenuating requirements, upgrades from one patch release of Rook to another are as simple as updating the common resources and the image of the Rook operator. For example, when Rook v1.21.1 is released, the process of updating from v1.21.0 is as simple as running the following:

1
2
git clone --single-branch --depth=1 --branch v1.21.1 https://github.com/rook/rook.git
cd rook/deploy/examples

If the Rook Operator or CephCluster are deployed into a different namespace than rook-ceph, see the Update common resources and CRDs section for instructions on how to change the default namespaces in common.yaml.

Then, apply the latest changes from v1.21, update CSI operator resources, and update the Rook Operator image.

1
2
kubectl apply -f common.yaml -f crds.yaml -f csi-operator.yaml
kubectl -n rook-ceph set image deploy/rook-ceph-operator rook-ceph-operator=rook/ceph:v1.21.1

A good practice is to update Rook common resources from the example manifests before any update. The common resources and CRDs might not be updated with every release, but Kubernetes will only apply updates to the ones that changed.

Also update optional resources like Prometheus monitoring noted more fully in the upgrade section below.

Helm

If Rook is installed via the Helm chart, Helm will handle some details of the upgrade itself. The upgrade steps in this guide will clarify what Helm handles automatically.

Important

Before upgrading to v1.20, ensure the cluster is already on at least v1.19.5. There was a critical update in v1.19.5 that will enable the helm upgrades to v1.20.

Upgrade charts in this order:

  1. rook-ceph
  2. ceph-csi-drivers (new requirement in v1.20)
  3. rook-ceph-cluster

Note

Be sure to update to a supported Helm version

rook-ceph Chart

The rook-ceph helm chart upgrade performs the Rook operator and ceph-csi-operator subchart upgrades.

To apply custom configuration to the ceph-csi-operator subchart, see the ceph-csi-operator configuration reference. Settings for the subchart need to be included in the ceph-csi-operator section of values.yaml when creating or updating the rook-ceph chart. See the default settings applied by Rook in values.yaml.

ceph-csi-drivers Chart

Install or upgrade the ceph-csi-drivers chart, to configure the CSI driver.

Important points to ensure CSI mounts continue working after upgrades:

  1. This ceph-csi-drivers chart must be installed, otherwise the CSI driver will be in a failed state due to missing service accounts.
  2. A recommended values.yaml is provided for Rook-compatible defaults in the drivers chart doc linked above. This is critical for setting the proper driver names with the rook operator namespace as a prefix.
  3. In previous releases, CSI settings were customized in the rook-ceph chart. Now see the CSI Drivers Chart config for customization.

rook-ceph-cluster Chart

The rook-ceph-cluster helm chart upgrade performs a Ceph upgrade if the Ceph image is updated.

Cluster Health

In order to successfully upgrade a Rook cluster, the following prerequisites must be met:

  • The cluster should be in a healthy state with full functionality. Review the health verification guide in order to verify a CephCluster is in a good starting state.
  • All pods consuming Rook storage should be created, running, and in a steady state.

Rook Operator Upgrade

The examples given in this guide upgrade a live Rook cluster running v1.19.11 to the version v1.21.0. This upgrade should work from any official patch release of Rook v1.20 to any official patch release of v1.21.

Let's get started!

Environment

These instructions will work for as long the environment is parameterized correctly. Set the following environment variables, which will be used throughout this document.

1
2
3
# Parameterize the environment
export ROOK_OPERATOR_NAMESPACE=rook-ceph
export ROOK_CLUSTER_NAMESPACE=rook-ceph

1. Update common resources and CRDs

Hint

Common resources and CRDs are automatically updated when using Helm charts.

First, apply updates to Rook common resources. This includes modified privileges (RBAC) needed by the Operator. Also update the Custom Resource Definitions (CRDs).

Get the latest common resources manifests that contain the latest changes.

1
2
git clone --single-branch --depth=1 --branch master https://github.com/rook/rook.git
cd rook/deploy/examples

If the Rook Operator or CephCluster are deployed into a different namespace than rook-ceph, update the common resource manifests to use your ROOK_OPERATOR_NAMESPACE and ROOK_CLUSTER_NAMESPACE using sed.

1
2
3
4
5
6
7
sed -i.bak \
    -e "s/\(.*\):.*# namespace:operator/\1: $ROOK_OPERATOR_NAMESPACE # namespace:operator/g" \
    -e "s/\(.*\):.*# namespace:cluster/\1: $ROOK_CLUSTER_NAMESPACE # namespace:cluster/g" \
    -e "s/\(.*serviceaccount\):.*:\(.*\) # serviceaccount:namespace:operator/\1:$ROOK_OPERATOR_NAMESPACE:\2 # serviceaccount:namespace:operator/g" \
    -e "s/\(.*serviceaccount\):.*:\(.*\) # serviceaccount:namespace:cluster/\1:$ROOK_CLUSTER_NAMESPACE:\2 # serviceaccount:namespace:cluster/g" \
    -e "s/\(.*\): [-_A-Za-z0-9]*\.\(.*\) # driver:namespace:cluster/\1: $ROOK_CLUSTER_NAMESPACE.\2 # driver:namespace:cluster/g" \
  common.yaml operator.yaml csi-operator.yaml cluster.yaml # add other files or change these as desired for your config

Apply the resources.

1
kubectl apply -f common.yaml -f crds.yaml -f csi-operator.yaml

Prometheus Updates

If Prometheus monitoring is enabled, upgrade the Prometheus RBAC resources:

1
kubectl apply -f deploy/examples/monitoring/rbac.yaml

2. Update the Rook Operator

Hint

The operator is automatically updated when using Helm charts.

The largest portion of the upgrade is triggered when the operator's image is updated to v1.21.x. When the operator is updated, it will proceed to update all of the Ceph daemons.

1
kubectl -n $ROOK_OPERATOR_NAMESPACE set image deploy/rook-ceph-operator rook-ceph-operator=rook/ceph:master

3. Update Ceph CSI Custom Images

Hint

This is automatically updated if custom CSI image versions are not set.

Update to the latest Ceph-CSI drivers if custom CSI images are specified. See the CSI Custom Images documentation.

4. Wait for the upgrade to complete

Watch now in amazement as the Ceph mons, mgrs, OSDs, rbd-mirrors, MDSes and RGWs are terminated and replaced with updated versions in sequence. The cluster may be unresponsive very briefly as mons update, and the Ceph Filesystem may fall offline a few times while the MDSes are upgrading. This is normal.

The versions of the components can be viewed as they are updated:

1
watch --exec kubectl -n $ROOK_CLUSTER_NAMESPACE get deployments -l rook_cluster=$ROOK_CLUSTER_NAMESPACE -o jsonpath='{range .items[*]}{.metadata.name}{"  \treq/upd/avl: "}{.spec.replicas}{"/"}{.status.updatedReplicas}{"/"}{.status.readyReplicas}{"  \trook-version="}{.metadata.labels.rook-version}{"\n"}{end}'

As an example, this cluster is midway through updating the OSDs. When all deployments report 1/1/1 availability and rook-version=v1.20.0, the Ceph cluster's core components are fully updated.

1
2
3
4
5
6
7
8
9
Every 2.0s: kubectl -n rook-ceph get deployment -o j...

rook-ceph-mgr-a         req/upd/avl: 1/1/1      rook-version=v1.21.0
rook-ceph-mon-a         req/upd/avl: 1/1/1      rook-version=v1.21.0
rook-ceph-mon-b         req/upd/avl: 1/1/1      rook-version=v1.21.0
rook-ceph-mon-c         req/upd/avl: 1/1/1      rook-version=v1.21.0
rook-ceph-osd-0         req/upd/avl: 1//        rook-version=v1.21.0
rook-ceph-osd-1         req/upd/avl: 1/1/1      rook-version=v1.20.7
rook-ceph-osd-2         req/upd/avl: 1/1/1      rook-version=v1.20.7

An easy check to see if the upgrade is totally finished is to check that there is only one rook-version reported across the cluster.

1
2
3
4
5
6
# kubectl -n $ROOK_CLUSTER_NAMESPACE get deployment -l rook_cluster=$ROOK_CLUSTER_NAMESPACE -o jsonpath='{range .items[*]}{"rook-version="}{.metadata.labels.rook-version}{"\n"}{end}' | sort | uniq
This cluster is not yet finished:
  rook-version=v1.20.7
  rook-version=v1.21.0
This cluster is finished:
  rook-version=v1.21.0

5. Verify the updated cluster**

At this point, the Rook operator should be running version rook/ceph:v1.21.0.

Verify the CSI drivers:

1
kubectl -n $ROOK_OPERATOR_NAMESPACE get operatorconfig,driver

Verify the CephCluster health using the health verification doc.