Does CA support auto-discovery of Azure VMSSs?

[sonnysideup]

The documentation in this project indicates that auto-discovery is only supported for aws and gce are supported, but the code for the Azure provider seems to indicate otherwise.

autoscaler/cluster-autoscaler/cloudprovider/azure/azure_manager.go

Lines 49 to 58 in cb4e60f

	type AzureManager struct {
	config *Config
	azClient *azClient
	env azure.Environment
	asgCache *asgCache
	lastRefresh time.Time
	asgAutoDiscoverySpecs []cloudprovider.LabelAutoDiscoveryConfig
	explicitlyConfigured map[string]bool
	}

autoscaler/cluster-autoscaler/cloudprovider/azure/azure_manager.go

Lines 265 to 268 in cb4e60f

	// Fetch automatically discovered ASGs. These ASGs should be unregistered if
	// they no longer exist in Azure.
	func (m *AzureManager) fetchAutoAsgs() error {
	groups, err := m.getFilteredAutoscalingGroups(m.asgAutoDiscoverySpecs)

Is it possible that the docs have been update since #542 was merged? I believe this should support VMSS in the same way it supports AWS ASGs now.

/cc @robinkraft

[MaciekPytel]

cc: @feiskyer

[feiskyer]

@sonnysideup It's an initial working version for vmss, e.g. by setting --node-group-auto-discovery=label:<key>=<value>, but still not fully supported yet. e.g. minSize and maxSize couldn't be set now.

[feiskyer]

/sig azure

[sonnysideup]

...but still not fully supported yet. e.g. minSize and maxSize couldn't be set now.

@feiskyer, thank you for your response. Just to make sure I understand this behavior, please let me frame an example and ask some questions:

Example Setup

1. CA is deployed into K8s (in Azure) with the setting --node-group-auto-discovery=label:foo=bar.
2. I then create a VMSS with the label foo=bar and set the min/max values on that resource during creation.

Questions

• If we can't set minSize nor maxSize, how will this behave when the CA starts managing my VMSS?
• If it's not fully-supported, what is currently supported?

[feiskyer]

It could discovery vmss based on labels, but setting minSize and maxSize is not supported yet.

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[fiunchinho]

Can this be closed now after merging #2121?

[feiskyer]

/close

[k8s-ci-robot]

@feiskyer: Closing this issue.

In response to this:

/close

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.