🐛 fix: set machine spec failuredomain string if not defined

[jcpowermac]

What this PR does / why we need it:

This commit sets the machine.Spec.FailureDomain string
from the vsphereMachine.Spec.FailureDomain if
defined and the machine failure domain is empty.

This resolves an issue where virtual machines are not placed
within vm-host groups in a timely
manner causing misplacement in the topology.

Fixes #3575

[k8s-ci-robot]

Welcome @jcpowermac!

It looks like this is your first PR to kubernetes-sigs/cluster-api-provider-vsphere 🎉. Please refer to our pull request process documentation to help your PR have a smooth ride to approval.

You will be prompted by a bot to use commands during the review process. Do not be afraid to follow the prompts! It is okay to experiment. Here is the bot commands documentation.

You can also check if kubernetes-sigs/cluster-api-provider-vsphere has its own contribution guidelines.

You may want to refer to our testing guide if you run into trouble with your tests not passing.

If you are having difficulty getting your pull request seen, please follow the recommended escalation practices. Also, for tips and tricks in the contribution process you may want to read the Kubernetes contributor cheat sheet. We want to make sure your contribution gets all the attention it needs!

Thank you, and welcome to Kubernetes. 😃

[k8s-ci-robot]

Hi @jcpowermac. Thanks for your PR.

I'm waiting for a kubernetes-sigs member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

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-sigs/prow repository.

[rvanderp3]

/ok-to-test

[rvanderp3]

/lgtm

/assign @sbueringer

[k8s-ci-robot]

LGTM label has been added.

Git tree hash: 29283286600b1e3b812f33127e495556fc7d4a05

[chrischdi]

Am I right that its always the way that:

Whatever is set on vSphereMachine's .spec.failureDomain gets propagated up (by CAPI's machine controller) to the Machine's .spec.failureDomain field.

Its just that the timing is bad (CAPI does this too late?!)

Otherwise: should we use a fallback to machine.Spec.FailureDomain, if its not set on vsphereMachine?

[chrischdi]

Reading the CAPI contract: https://main.cluster-api.sigs.k8s.io/developer/providers/contracts/infra-machine#inframachine-failure-domain

It should be the other way around (which means code as-is should be correct?!).

[jcpowermac]

used rr to troubleshoot, the machine object doesn't receive the failuredomain name before the guest is powered on - then the race is on...will the node ccm init prior to being added to the vm-host group and potentially vmotion'ed to the correct location in the vCenter cluster.

cluster-api-provider-vsphere/pkg/services/govmomi/service.go

Lines 183 to 195 in ade4688

	if ok, err := vms.reconcileVMGroupInfo(ctx, virtualMachineCtx); err != nil || !ok {
	return vm, err
	}
	if err := vms.reconcileClusterModuleMembership(ctx, virtualMachineCtx); err != nil {
	return vm, err
	}
	if ok, err := vms.reconcilePowerState(ctx, virtualMachineCtx); err != nil || !ok {
	return vm, err
	}
	if err := vms.reconcileHostInfo(ctx, virtualMachineCtx); err != nil {

cluster-api-provider-vsphere/pkg/services/govmomi/service.go

Line 726 in ade4688

if virtualMachineCtx.VSphereFailureDomain == nil || virtualMachineCtx.VSphereFailureDomain.Spec.Topology.Hosts == nil {

[chrischdi]

I agree that potentially is the behavior when setting failuredomain only on the InfraMachine object (VSphereMachineTemplate?!).

According to kubernetes-sigs/cluster-api#11232 the correct usage is setting FailureDomain on the Machine object instead. Which means it would also be there in the referenced code.

Setting it on the VSphereMachine / VSphereMachineTemplate is the old deprecated behavior.

[jcpowermac]

I understand that its deprecated behavior, what is the solution then? The machine failuredomain name is slow to become available in the reconcile loop, which causes this race.

[chrischdi]

The solution if I'm right is: Set the failuredomain on the Machine (or MachineSet or MachineDeployment) object instead of the InfraMachine / VSphereMachine.

It should be fine to still set it on the InfraMachine / VSphereMachine. The important piece is to set it on the Machine instead of relying on CAPI copy this information to the InfraMachine / VSphereMachine.

[k8s-ci-robot]

New changes are detected. LGTM label has been removed.

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please ask for approval from sbueringer. For more information see the Code Review Process.

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

Needs approval from an approver in each of these files:

• OWNERS

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