Bug 1982868: UPSTREAM: <carry>: admission/managementcpusoverride: cover the roll-back case #877

cynepco3hahue · 2021-08-03T17:32:29Z

During the upgrade and roll-back flow 4.7->4.8->4.7, the topology related
fields under the infrastructure can be empty because the
old API does not support them.

The code will equal the empty infrastructure section with the current one.
When the status has some other non-empty field, and topology fields
are empty, we assume that the cluster currently passes
via roll-back and not via the clean install.

Signed-off-by: Artyom Lukianov alukiano@redhat.com

openshift-ci · 2021-08-03T17:32:32Z

@cynepco3hahue: This pull request references Bugzilla bug 1982868, which is invalid:

expected the bug to be open, but it isn't
expected the bug to target the "4.9.0" release, but it targets "4.8.0" instead
expected the bug to be in one of the following states: NEW, ASSIGNED, ON_DEV, POST, POST, but it is CLOSED (DUPLICATE) instead

Comment /bugzilla refresh to re-evaluate validity if changes to the Bugzilla bug are made, or edit the title of this pull request to link to a different bug.

In response to this:

Bug 1982868: UPSTREAM: : cover the roll-back case

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

openshift-ci-robot · 2021-08-03T17:32:37Z

@cynepco3hahue: the contents of this pull request could not be automatically validated.

The following commits could not be validated and must be approved by a top-level approver:

1e44679|UPSTREAM: : cover the roll-back case: does not specify an upstream backport in the commit message

cynepco3hahue · 2021-08-03T17:33:51Z

/bugzilla refresh

openshift-ci · 2021-08-03T17:34:59Z

@cynepco3hahue: This pull request references Bugzilla bug 1982868, which is valid. The bug has been moved to the POST state. The bug has been updated to refer to the pull request using the external bug tracker.

3 validation(s) were run on this bug

bug is open, matching expected state (open)
bug target release (4.9.0) matches configured target release for branch (4.9.0)
bug is in the state ASSIGNED, which is one of the valid states (NEW, ASSIGNED, ON_DEV, POST, POST)

No GitHub users were found matching the public email listed for the QA contact in Bugzilla (schoudha@redhat.com), skipping review request.

In response to this:

/bugzilla refresh

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

wking · 2021-08-03T17:43:32Z

openshift-kube-apiserver/admission/autoscaling/managementcpusoverride/admission.go

+	// the infrastructure status is not empty, but topology related fields do not have any values indicates that
+	// the cluster is during the roll-back process to the version that does not support the topology fields
+	// the upgrade to 4.8 handled by the CR defaulting
+	if clusterInfra.Status.ControlPlaneTopology == "" && clusterInfra.Status.InfrastructureTopology == "" {


you don't actually need this block, right? Because if they are empty strings for whatever reason, the != configv1.SingleReplicaTopologyMode escape immediately below will already return nil.

I prefer an explicit statement rather than a coincidence in code that does not intentionally handle this case.

sttts · 2021-08-03T17:50:17Z

Please use commit titles that make sense out of context. This is about management cpu separation. So put something like admission/managementcpusoverride: into the commit message (and PR title). Makes our lifes easier later.

sttts · 2021-08-03T17:53:35Z

openshift-kube-apiserver/admission/autoscaling/managementcpusoverride/admission.go

-	if clusterInfra.Status.ControlPlaneTopology == "" || clusterInfra.Status.InfrastructureTopology == "" {
-		return admission.NewForbidden(attr, fmt.Errorf("%s infrastructure resource has empty status.controlPlaneTopology or status.infrastructureTopology", PluginName))
+	// the infrastructure status is empty, so we can not decide the cluster type
+	if reflect.DeepEqual(clusterInfra.Status, configv1.InfrastructureStatus{}) {


~~This breaks when somebody adds a default to a status field. You might decide that this is fine, but this needs a smoke test that make sure we would notice. This is a blocker for the PR.~~

This works even with defaults of status, as long as there is a field in InfrastructureStatus that is set whenever the status is written, and this is always the case, right? So I think we only need a clear godoc comment why this works. Something along ^ lines.

openshift-ci-robot · 2021-08-04T14:08:16Z

@cynepco3hahue: the contents of this pull request could not be automatically validated.

The following commits could not be validated and must be approved by a top-level approver:

3d00a0c|UPSTREAM: : cadmission/managementcpusoverride: cover the roll-back case: does not specify an upstream backport in the commit message

…ack case During the upgrade and roll-back flow 4.7->4.8->4.7, the topology related fields under the infrastructure can be empty because the old API does not support them. The code will equal the empty infrastructure section with the current one. When the status has some other non-empty field, and topology fields are empty, we assume that the cluster currently passes via roll-back and not via the clean install. Signed-off-by: Artyom Lukianov <alukiano@redhat.com>

openshift-ci-robot · 2021-08-04T14:21:15Z

@cynepco3hahue: the contents of this pull request could not be automatically validated.

The following commits could not be validated and must be approved by a top-level approver:

3d00a0c|UPSTREAM: : cadmission/managementcpusoverride: cover the roll-back case: does not specify an upstream backport in the commit message

sttts · 2021-08-04T14:58:13Z

/lgtm
/approve

openshift-ci · 2021-08-04T15:03:28Z

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: cynepco3hahue, sttts

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

~~DOWNSTREAM_OWNERS~~ [sttts]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

cynepco3hahue · 2021-08-04T16:21:32Z

/hold
@sttts @soltysh Should it be merged only after the rebase?

cynepco3hahue · 2021-08-08T15:47:18Z

/retest

openshift-bot · 2021-08-18T02:57:55Z

/retest-required