AzureCluster failureDomains do not always match specific control plane AzureMachine SKU

[nojnhuh]

/kind bug

What steps did you take and what happened:

CAPZ's AzureCluster controller contains logic to set the status.failureDomains to the available availability zones for the region of the cluster resources:

cluster-api-provider-azure/azure/services/resourceskus/cache.go

Line 170 in 89efbf3

func (c *Cache) GetZones(ctx context.Context, location string) ([]string, error) {

This function loops through all VM SKUs, returning any availability zone without restrictions for at least one SKU to be set in the AzureCluster's failureDomains. But when a specific VM SKU in a region has a Zone restriction preventing VMs from being created in specific zones, CAPZ may still list those zones in its status.failureDomains if some other SKU in the same region does not have the same restrictions, causing VMs to be created in zones with restrictions and fail.

What did you expect to happen:

The AzureCluster failureDomains match the availability of the specific SKU of control plane VM being used.

Anything else you would like to add:
[Miscellaneous information that will assist in solving the issue.]

Environment:

• cluster-api-provider-azure version:
• Kubernetes version: (use kubectl version):
• OS (e.g. from /etc/os-release):

[mboersma]

@nojnhuh should this be in the next milestone?

[nojnhuh]

I don't think this is something we need to do soon, at least since we've ironed out the regions we use in CI to avoid this.

[willie-yao]

/priority backlog

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[nojnhuh]

/remove-lifecycle stale