# Kuberay operator upgrade from v1.0.0 to v1.2.2

**URL:** <https://discuss.ray.io/t/kuberay-operator-upgrade-from-v1-0-0-to-v1-2-2/16019>\
**Category:** Kubernetes\
**Created:** [October 16, 2024, 12:40pm UTC](https://discuss.ray.io/t/kuberay-operator-upgrade-from-v1-0-0-to-v1-2-2/16019 "2024-10-16T12:40:18Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Sowmith\_Renumakala](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/sowmith_renumakala/32/6646_2.png) [@Sowmith\_Renumakala](https://discuss.ray.io/u/Sowmith_Renumakala)\
**Post date:** [October 16, 2024, 12:40pm UTC](https://discuss.ray.io/t/kuberay-operator-upgrade-from-v1-0-0-to-v1-2-2/16019/1 "2024-10-16T12:40:18Z")

</div>

Hello

We have kuberay operator which is deployed at namespace level at version v1.0.0. We want to upgrade to latest version of v1.2.2. Cant find enough documentation on

- If there are any breaking changes of directly upgrading to v1.2.2 from v1.0.0 or the prerequisites
- I see that for version v1.1.0 ray version needs to be 2.8.0 or higher. What happens to existing jobs or services once we upgrade to latest version 1.2.2, would they start erroring out ?
- Whats clean way to upgrade

Thanks & Regards

---

<div class="post-metadata">

**Author:** ![Chi-Sheng\_Liu](https://sea2.discourse-cdn.com/flex020/user_avatar/discuss.ray.io/chi-sheng_liu/32/6627_2.png) [@Chi-Sheng\_Liu](https://discuss.ray.io/u/Chi-Sheng_Liu)\
**Post date:** [October 18, 2024, 10:30am UTC](https://discuss.ray.io/t/kuberay-operator-upgrade-from-v1-0-0-to-v1-2-2/16019/2 "2024-10-18T10:30:04Z")

</div>

Please refer to the [KubeRay upgrade guide](https://docs.ray.io/en/master/cluster/kubernetes/user-guides/upgrade-guide.html).

Upgrading from KubeRay v1.0.0 to v1.2.2 requires first upgrading the CRD:

```bash
kubectl replace -k "github.com/ray-project/kuberay/ray-operator/config/crd?ref=v1.2.2"

```

If there are still v1alpha Ray resources in your cluster, you will need to use `kubectl edit` or `kubectl apply` to update them to v1.

Additionally, if any Ray resources are using a Ray version below 2.8.0, you need to upgrade them to 2.8.0 or above.

Note that changing the image in the RayCluster spec (e.g., from `rayproject/ray:2.0.0` to `rayproject/ray:2.9.0`) will not automatically recreate the existing pods associated with this RayCluster. The existing pods will not restart automatically; instead, only when the old pods are deleted will new pods be created with the updated image. So for RayCluster you need to either delete and recreate it or use `suspend` in its spec to delete all pods. For RayService, changing the image in the RayService spec will trigger a zero-downtime upgrade, so a new RayCluster will be created for you.

Finally, you can upgrade the KubeRay operator itself:

```bash
helm upgrade kuberay-operator kuberay/kuberay-operator --version v1.2.2

```