Version: openshift-ansible-4.10.0-202112151834.p0.gcc445ce.assembly.stream.el7.noarch.rpm What happened? As our official doc clarified the RHEL-7 workers supposed to be removed in 4.9[1] "You cannot upgrade RHEL 7 compute machines to RHEL 8. You must deploy new RHEL 8 hosts, and the old RHEL 7 hosts should be removed." but there's still some cases that customer would try to upgrade their RHEL-7 workers to 4.10 from 4.9 or 4.8(EUS-to-EUS upgrade). So it's better to add a check in ansible code before we start the upgrade steps against RHEL-7 workers. [1] https://docs.openshift.com/container-platform/4.9/updating/updating-cluster-rhel-compute.html#rhel-compute-updating-minor_updating-cluster-rhel-compute What did you expect to happen? upgrade playbook should abort directly when detecting it's upgrading a RHEL-7 worker, and prompt meaningful message like: RHEL-7 worker are NOT supported in 4.10, you must deploy new RHEL-8 or RHCOS workers, the old RHEL 7 hosts should be removed.
This is a good suggestion and we should incorporate it, but trying to determine if this is a blocker: - If we don't make this change it would allow someone to upgrade into an unsupported state, but it is also documented that it is unsupported. - openshift-ansible is not shipped as part of the release image, so how does that relate to the entire blocker+/- system?
Given that openshift-ansible is not part of the release, this by definition cannot be a blocker.
(In reply to Matthew Staebler from comment #2) > Given that openshift-ansible is not part of the release, this by definition > cannot be a blocker. Actually, that is not quite right. I could see a scenario where something outside of the release is required to be in place or fixed prior to the release. I could also see where the openshift-ansible playbooks would fit that description. Nevertheless, I do not think this is meets the bar for blocking the 4.10 release.
In case my opinion matters, I agree that this should not be considered a blocker.
Verified this bug with openshift-ansible-4.11.0-202202111945.p0.g4f59ed0.assembly.stream.el7.noarch.rpm For RHEL-7 host, when trying to run playbooks/upgrade.yml or playbooks/scaleup.yml, it would fail with the following error. TASK [openshift_node : Set fact l_cluster_version] ***************************** Saturday 12 February 2022 13:39:22 +0800 (0:00:00.693) 0:00:02.027 ***** ok: [ip-10-0-51-150.us-east-2.compute.internal] => {"ansible_facts": {"l_cluster_version": "4.11"}, "changed": false} TASK [openshift_node : Fail if not using RHEL8 beginning with version 4.10] **** Saturday 12 February 2022 13:39:22 +0800 (0:00:00.088) 0:00:02.116 ***** fatal: [ip-10-0-51-150.us-east-2.compute.internal]: FAILED! => {"changed": false, "msg": "As of v4.10, RHEL nodes must be at least version 8.4"}
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Important: OpenShift Container Platform 4.11.0 bug fix and security update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2022:5069