Red Hat OpenShift Service on AWS (ROSA) with Hosted Control Planes (HCP): Transitioning from Technology Preview to Fully Supported Clusters
Thank you for trying Red Hat OpenShift Service on AWS (ROSA) with hosted control planes (HCP) during our Technology Preview period. We hope this experience has exceeded your expectations and invite you to transition to fully supported ROSA HCP clusters backed by our 99.95% uptime SLA.
End of Technology Preview Period
Any clusters provisioned before 4 December 2023 UTC will have been provisioned as a Technology Preview cluster. Clusters provisioned after the 4 December 2023 UTC will be provisioned as a fully supported Managed Service.
Tech Preview Deprovisioning
The tech preview clusters will continue to function through December 2023. After this date, the clusters will start to be deprovisioned through January2024. Deprovisioning refers to the removal of control plane components managed by Red Hat and to the deletion of the HCP resources deployed in the customer's AWS account. This means AWS resources managed by Red Hat HCP but located in customer AWS accounts will be impacted.
NOTE: It is recommended that any testing using the Tech Preview clusters be completed by the end of December 2023, along with collecting any resource configurations in AWS and HCP that you may wish to refer to in the future, as this will be lost as part of the deprovisioning.
Warnings of the deprovisioning actions were first provided as part of the hosted control plane install (HCP;--hosted-cp) and were followed up with a warning message in the Cluster History tab available in the Hybrid Cloud Console for the HCP cluster.
Manual Deprovisioning
Deprovisioning both types of HCP clusters (Fully support or Technical preview) can be done manually at any time using the following rosa CLI command.
rosa delete cluster --cluster <ClusterName|Id>
NOTE: Deleting the AWS resources from the cluster owner's AWS account does not deprovisioning the cluster as there are still AWS and other resources fully managed by SRE that are inaccessible to the cluster owner.
AWS Artifacts
Under certain conditions, some AWS resources (artefacts) could remain in the cluster owner's AWS account.
Inaccessible AWS Accounts
If the AWS roles used by Red Hat to access the HCP resources have been modified or revoked, then the deprovisioning may fail. If this occurs, some cluster resources may be left running in the AWS account owned by the cluster owner and will require manual deletion by the cluster owner. Not doing so could lead to cluster owners being billed by AWS.
Partially Deleted AWS Resources
If the cluster owner has deleted some of the HCP resources, then the deprovisoning process may fail. If this occurs, some cluster resources may be left running in the AWS account owned by the cluster owner and will require manual deletion by the cluster owner. Not doing so could lead to cluster owners being billed by AWS.
Additional Resources
Deprovisoning will attempt to delete all the HCP managed resources in the cluster owner AWS account, but any additional resources are out of scope and will remain. These additional resources are the responsibility of the cluster owner to remove. Not doing so could lead to cluster owners being billed by AWS.
Important Dates
- 4 January 2024: Deadline for users to de-provision their ROSA with HCP Technology Preview clusters.
- Post 4 January 2024: Red Hat will begin automatically deprovisioning any remaining Technology Preview clusters.
Recommended Next Steps
- Review Public Announcement: For a detailed understanding of the benefits and architectural improvements in ROSA with HCP, refer to our public announcement.
- Plan Your Transition: Assess your current cluster setup, collect any non-default configurations and plan the transition to a fully supported cluster, considering application dependencies and potential downtime.
- De-provision Technology Preview Clusters: Ensure that you de-provision your Technology Preview clusters by 4 January 2024 to avoid automatic de-provisioning.
Comments