> ## Documentation Index
> Fetch the complete documentation index at: https://docs.domino.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Google Kubernetes Engine (GKE) upgrade guide

This topic describes how to upgrade Kubernetes in your [Google Kubernetes Engine (GKE)](https://cloud.google.com/kubernetes-engine/docs/concepts/kubernetes-engine-overview) Domino deployment. GKE is hosted on [Google Cloud Platform (GCP)](https://cloud.google.com/docs/overview).

<Warning>
  Immediately after the Kubernetes upgrade, you must [upgrade Domino](/6.3/admin/operations/upgrade-domino) to a [compatible version of Kubernetes](/6.3/admin/architecture-and-deployment/kubernetes-infrastructure/kubernetes-compatibility). Domino will not work as expected until this is completed.

  For example, after upgrading Kubernetes to v1.22, you must upgrade to Domino v5.2 or later. Kubernetes v1.22 is not compatible with older versions of Domino. Similarly, after upgrading to Kubernetes 1.23 or 1.24, you must upgrade to Domino v5.3 or later.
</Warning>

## Prerequisites

To upgrade Kubernetes in GKE, you must have a Unix or Linux terminal with the following:

* [gcloud CLI](https://cloud.google.com/sdk/docs/install?hl=fr#deb) installed.

* [Kubectl](https://kubernetes.io/docs/tasks/tools/install-kubectl-linux/) installed.

Use environment variables to set the values of IDs, names, and labels. This simplifies the commands you’ll run while upgrading:

```shell theme={null}
export CLUSTER_PROJECT=<The project name of the cluster>
export CLUSTER_NAME=<The name of the cluster>
export CLUSTER_REGION=<The region where the cluster is located>
```

## Get cluster credentials

1. If you aren’t already signed in with gcloud, run `gcloud init`.

2. To protect your current configuration file from being overwritten, run `export KUBECONFIG=$(pwd)/kubeconfig`.

3. Get the remote `kubeconfig`:

   ```shell theme={null}
   gcloud container clusters get-credentials $CLUSTER_NAME --region $CLUSTER_REGION --project $CLUSTER_PROJECT
   ```

## Disable the disruption budgets

You must temporarily disable the `PodDisruptionBudget` (PDB) for certain services. To do this, back up the PDBs, delete the original PDBs, then update Kubernetes and restore the PDBs from the backup after the update is complete.

1. Back up the PDBs:

   ```shell theme={null}
   kubectl get pdb -n domino-platform -o yaml mongodb-arbiter > mongodb-arbiter.yml
   kubectl get pdb -n domino-platform -o yaml mongodb-primary > mongodb-primary.yml
   kubectl get pdb -n domino-platform -o yaml mongodb-secondary > mongodb-secondary.yml
   ```

2. Delete the PDBs:

   ```shell theme={null}
   kubectl delete pdb -n domino-platform mongodb-arbiter
   kubectl delete pdb -n domino-platform mongodb-primary
   kubectl delete pdb -n domino-platform mongodb-secondary
   ```

## Determine the upgrade version

1. Get the available GKE versions:

   ```shell theme={null}
   gcloud container get-server-config \
     --region=${CLUSTER_REGION} \
     --flatten="channels" \
     --format="yaml(channels.channel,channels.validVersions)"
   ```

2. Get the current master version:

   ```shell theme={null}
   gcloud container clusters describe ${CLUSTER_NAME} --region=${CLUSTER_REGION} --format="value(currentMasterVersion)"
   ```

<Note>
  You can only upgrade by one minor version at a time. For example, if you are on Kubernetes version 1.22, you can upgrade to version 1.23.
</Note>

## Update the cluster

1. Upgrade the control plane. This might take several minutes to complete. Monitor the GCP console for the status of the cluster upgrade.

   ```shell theme={null}
   gcloud container clusters upgrade $CLUSTER_NAME --project $CLUSTER_PROJECT --master --region $CLUSTER_REGION --cluster-version <version-number>
   ```

2. Upgrade the node pools:

   ```shell theme={null}
   gcloud container clusters upgrade $CLUSTER_NAME --project $CLUSTER_PROJECT --region $CLUSTER_REGION --node-pool=platform
   gcloud container clusters upgrade $CLUSTER_NAME --project $CLUSTER_PROJECT --region $CLUSTER_REGION --node-pool=compute
   gcloud container clusters upgrade $CLUSTER_NAME --project $CLUSTER_PROJECT --region $CLUSTER_REGION --node-pool=gpu
   ```

## Restore the disruption budgets

Run these commands from the same folder as the backup:

```shell theme={null}
kubectl apply -f mongodb-arbiter.yml
kubectl apply -f mongodb-primary.yml
kubectl apply -f mongodb-secondary.yml
```