instance

package
v1.7.0-test.batchlistD... Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Mar 18, 2026 License: Apache-2.0, Apache-2.0 Imports: 52 Imported by: 0

Documentation

Overview

This file is the central registry for fields that receive special treatment during batch VM creation. It lives in the instance package (next to buildAKSMachineTemplate) so that developers modifying the template builder naturally encounter it when adding or changing per-NodeClaim fields.

The batch system (pkg/providers/instance/batch/) imports these functions to decide which fields to exclude from the grouping hash and which to strip from the shared request body.

## When to update this file

Update ClearPerMachineFields if you add a MachineProperties field whose value varies per NodeClaim (like Tags, which carry the NodeClaim name). Fields that are the same for all NodeClaims sharing a NodePool+NodeClass (like VMSize, Priority, KubeletConfig) should NOT be added here — they are part of the shared template and must influence the batch grouping hash.

Update ClearReadOnlyFields if the Azure SDK adds a new server-set field to MachineProperties that should never influence hashing or request bodies.

## Design context (PR #1455 review discussion)

This registry was introduced to solve a developer-experience gap: when someone modifies buildAKSMachineTemplate() to set a new per-NodeClaim field, they need to also exclude it from the batch grouping hash. Without this file in the same package, that requirement was invisible — the batch module is a subpackage that template-builder developers may never look at.

The reflection-based guardrail test (TestComputeTemplateHash_AllFieldsAccountedFor in batch/grouper_test.go) catches new SDK fields at test time. This registry catches new per-NodeClaim field wiring at development time by being co-located with the template builder.

## Known gap: offerings-related fields

Some fields like Machine.Zones are selected from instance type offerings within the instance package, not from NodeClaim-unique data. They are per-machine but not per-NodeClaim — they come from the scheduler's zone selection. These live on the Machine struct (not MachineProperties), so they don't go through this registry. If the Machine API adds new offerings-related fields to MachineProperties in the future, this distinction will need to be revisited.

## Future improvement: type-safe config split (Proposal B)

A stronger guarantee would be to split the machine template into typed structs (SharedMachineConfig vs PerMachineConfig) at the point of creation, making the type system enforce the separation. See PR #XXXX for that proposal. Discussion: https://github.com/Azure/karpenter-provider-azure/pull/1455#discussion_r2901097563

Index

Constants

This section is empty.

Variables

View Source 
var (
	KarpCapacityTypeToAKSScaleSetPriority = map[string]armcontainerservice.ScaleSetPriority{
		karpv1.CapacityTypeSpot:     armcontainerservice.ScaleSetPrioritySpot,
		karpv1.CapacityTypeOnDemand: armcontainerservice.ScaleSetPriorityRegular,
	}
	AKSScaleSetPriorityToKarpCapacityType = map[armcontainerservice.ScaleSetPriority]string{
		armcontainerservice.ScaleSetPrioritySpot:    karpv1.CapacityTypeSpot,
		armcontainerservice.ScaleSetPriorityRegular: karpv1.CapacityTypeOnDemand,
	}
)
View Source 
var (
	// VMCreateStartMetric tracks when VM creation starts.
	//
	// STABILITY: ALPHA - This metric may change or be removed without notice.
	VMCreateStartMetric = prometheus.NewCounterVec(
		prometheus.CounterOpts{
			Namespace: metrics.Namespace,
			Subsystem: instanceSubsystem,
			Name:      "vm_create_start_total",
			Help:      "Total number of VM creation operations started.",
		},
		[]string{metrics.ImageLabel, metrics.SizeLabel, metrics.ZoneLabel, metrics.CapacityTypeLabel, metrics.NodePoolLabel},
	)

	// VMCreateFailureMetric tracks VM creation failures, regardless of phase.
	//
	// STABILITY: ALPHA - This metric may change or be removed without notice.
	VMCreateFailureMetric = prometheus.NewCounterVec(
		prometheus.CounterOpts{
			Namespace: metrics.Namespace,
			Subsystem: instanceSubsystem,
			Name:      "vm_create_failure_total",
			Help:      "Total number of VM creation failures.",
		},
		[]string{metrics.ImageLabel, metrics.SizeLabel, metrics.ZoneLabel, metrics.CapacityTypeLabel, metrics.NodePoolLabel, metrics.PhaseLabel, metrics.ErrorCodeLabel},
	)
)

We don't need to add disk specification since they are statically defined and can be traced with provided labels.

View Source 
var (
	KarpCapacityTypeToVMPriority = map[string]armcompute.VirtualMachinePriorityTypes{
		karpv1.CapacityTypeSpot:     armcompute.VirtualMachinePriorityTypesSpot,
		karpv1.CapacityTypeOnDemand: armcompute.VirtualMachinePriorityTypesRegular,
	}
	VMPriorityToKarpCapacityType = map[armcompute.VirtualMachinePriorityTypes]string{
		armcompute.VirtualMachinePriorityTypesSpot:    karpv1.CapacityTypeSpot,
		armcompute.VirtualMachinePriorityTypesRegular: karpv1.CapacityTypeOnDemand,
	}
	// Note that there is no ScaleSetPriorityToKarpCapacityType because the karpenter.sh/capacity-type
	// label is the "official" label that we actually key priority off of. Selection still works though
	// because when we list instance types on-demand offerings always have v1beta1.ScaleSetPriorityRegular
	// and spot instances always have v1beta1.ScaleSetPrioritySpot, so the correct karpenter.sh/capacity-type
	// label is still selected even if the user is using kubernetes.azure.com/scalesetpriority only on the NodePool.
	VMPriorityToScaleSetPriority = map[armcompute.VirtualMachinePriorityTypes]string{
		armcompute.VirtualMachinePriorityTypesSpot:    v1beta1.ScaleSetPrioritySpot,
		armcompute.VirtualMachinePriorityTypesRegular: v1beta1.ScaleSetPriorityRegular,
	}
)
View Source 
var (
	NodePoolTagKey = strings.ReplaceAll(karpv1.NodePoolLabelKey, "/", "_")
)

Functions

func AKSMachineTimestampFromTag added in v1.7.0 

func AKSMachineTimestampFromTag(timestampStr string) (time.Time, error)

AKSMachineTimestampFromTag parses an AKS machine creation timestamp tag value back to time.Time

func AKSMachineTimestampToMeta added in v1.7.0 

func AKSMachineTimestampToMeta(t time.Time) metav1.Time

AKSMachineTimestampToMeta converts a time.Time to metav1.Time for AKS machine creation timestamps

func AKSMachineTimestampToTag added in v1.7.0 

func AKSMachineTimestampToTag(t time.Time) string

AKSMachineTimestampToTag converts a time.Time to the string format used in AKS machine creation timestamp tags

func BuildJSONFromAKSMachine added in v1.7.0 

func BuildJSONFromAKSMachine(aksMachine *armcontainerservice.Machine) string

BuildJSONFromAKSMachine returns a JSON string representation of an AKS Machine for logging/debugging purposes. Returns an error string if marshaling fails.

func BuildNodeClaimFromAKSMachine added in v1.7.0 

func BuildNodeClaimFromAKSMachine(ctx context.Context, aksMachine *armcontainerservice.Machine, possibleInstanceTypes []*corecloudprovider.InstanceType, aksMachineLocation string) (*karpv1.NodeClaim, error)

Expect AKS machine struct to be fully populated as if it comes from GET. Not assuming that NodeClaim exists.

func BuildNodeClaimFromAKSMachineTemplate added in v1.7.0 

func BuildNodeClaimFromAKSMachineTemplate(
	ctx context.Context,
	aksMachineTemplate *armcontainerservice.Machine,
	instanceType *corecloudprovider.InstanceType,
	capacityType string,
	zone *string,
	aksMachineResourceID string,
	vmResourceID string,
	isDeleting bool,
	aksMachineNodeImageVersion string,
) (*karpv1.NodeClaim, error)

Note that the template is not guaranteed to have status fields, thus they are made explicit here. Other Karpenter-level fields are also included as they may be easily retrieved during templating phase. Not assuming that NodeClaim exists.

func ClearPerMachineFields 

func ClearPerMachineFields(props *armcontainerservice.MachineProperties)

ClearPerMachineFields zeros MachineProperties fields that vary per machine and travel via the BatchPutMachine HTTP header instead of the shared request body.

This is the single source of truth for per-machine MachineProperties fields. Both the batch grouper (computeTemplateHash) and coordinator (ExecuteBatch) call this function, so adding a field here automatically:

  • Excludes it from the batch grouping hash (machines differing only on this field will batch together)
  • Strips it from the shared API request body (the field travels in the BatchPutMachine header instead)

When adding a field here, also:

  1. Update the excludedFields map in TestComputeTemplateHash_AllFieldsAccountedFor (batch/grouper_test.go)
  2. Add extraction logic in Coordinator.buildBatchHeader (batch/coordinator.go) so the field value is included in the per-machine header entries
  3. Add the field to MachineEntry in batch/types.go

func ClearReadOnlyFields 

func ClearReadOnlyFields(props *armcontainerservice.MachineProperties)

ClearReadOnlyFields zeros MachineProperties fields that are set by the server and should never influence template hashing or request bodies.

These fields appear on MachineProperties in GET responses but are not part of the creation template. Adding a new server-set field here also requires updating the excludedFields map in TestComputeTemplateHash_AllFieldsAccountedFor.

func ConfigureAKSMachineTags added in v1.7.0 

func ConfigureAKSMachineTags(opts *options.Options, nodeClass *v1beta1.AKSNodeClass, nodeClaim *karpv1.NodeClaim, creationTimestamp time.Time) map[string]*string

ConfigureAKSMachineTags returns the tags to be applied to AKS machine instances and their affiliated resources. This includes all standard tags plus the AKS machine distinguishing tag.

func ConvertToVirtualMachineIdentity 

func ConvertToVirtualMachineIdentity(nodeIdentities []string) *armcompute.VirtualMachineIdentity

func CreateVirtualMachine 

func CreateVirtualMachine(ctx context.Context, client azclient.VirtualMachinesAPI, rg, vmName string, vm armcompute.VirtualMachine) (*armcompute.VirtualMachine, error)

func ErrorCodeForMetrics added in v1.6.6 

func ErrorCodeForMetrics(err error) string

ErrorCodeForMetrics extracts a stable Azure error code for metric labeling when possible.

func FindNodePoolFromAKSMachine added in v1.7.0 

func FindNodePoolFromAKSMachine(ctx context.Context, aksMachine *armcontainerservice.Machine, kubeClient client.Client) (*karpv1.NodePool, error)

May return apimachinery.NotFoundError if NodePool is not found.

func GenerateResourceName 

func GenerateResourceName(nodeClaimName string) string

E.g., aks-default-2jf98

func GetAKSLabelZoneFromAKSMachine added in v1.7.0 

func GetAKSLabelZoneFromAKSMachine(aksMachine *armcontainerservice.Machine, location string) (string, error)

GetAKSLabelZoneFromAKSMachine returns the zone for the given AKS machine, or an empty string if there is no zone specified This function is analogous to utils.GetAKSLabelZoneFromVM but for AKS machines

func GetAKSMachineNameFromNodeClaim added in v1.7.0 

func GetAKSMachineNameFromNodeClaim(nodeClaim *karpv1.NodeClaim) (string, bool)

GetAKSMachineNameFromNodeClaim extracts the AKS machine name from the NodeClaim annotations Returns false if the annotation is not present or the value is empty, which can indicate that the NodeClaim is not associated with an AKS machine.

func GetAKSMachineNameFromNodeClaimName added in v1.7.0 

func GetAKSMachineNameFromNodeClaimName(nodeClaimName string) (string, error)

ASSUMPTION: NodeClaim name is in the format of <NodePool name>-<hash suffix> If total length exceeds AKS machine name limit, the exceeded part will be replaced with another deterministic hash. E.g., "thisisalongnodepoolname-a1b2c" --> "thisisalongnoz9y8x7-a1b2c"

func GetAKSMachineNameFromVMName added in v1.7.0 

func GetAKSMachineNameFromVMName(aksMachinesPoolName, vmName string) (string, error)

vmName = aks-<machinesPoolName>-<aksMachineName>-########-vm This is distinguishable from VM instance name as its suffix will always be 5 alphanumerics rather than "vm"

func GetCapacityTypeFromVM added in v1.6.6 

func GetCapacityTypeFromVM(vm *armcompute.VirtualMachine) string

func GetManagedExtensionNames added in v1.6.2 

func GetManagedExtensionNames(provisionMode string, env *auth.Environment) []string

GetManagedExtensionNames gets the names of the VM extensions managed by Karpenter. This is a set of 1 or 2 extensions (depending on provisionMode): aksIdentifyingExtension and (sometimes) cse.

func GetNICListQueryBuilder added in v0.6.2 

func GetNICListQueryBuilder(rg string) *kql.Builder

GetNICListQueryBuilder returns a KQL query builder for listing NICs with nodepool tags

func GetScaleSetPriorityLabelFromVM added in v1.7.0 

func GetScaleSetPriorityLabelFromVM(vm *armcompute.VirtualMachine) string

func GetVMListQueryBuilder added in v0.6.2 

func GetVMListQueryBuilder(rg string) *kql.Builder

GetVMListQueryBuilder returns a KQL query builder for listing VMs with nodepool tags

func IsAKSMachineOrMachinesPoolNotFound added in v1.6.8 

func IsAKSMachineOrMachinesPoolNotFound(err error) bool

func NewAKSMachineTimestamp added in v1.7.0 

func NewAKSMachineTimestamp() time.Time

NewAKSMachineTimestamp returns the current time truncated to centisecond precision for AKS machine creation timestamps

func NewQueryRequest 

func NewQueryRequest(subscriptionID *string, query string) *arg.QueryRequest

func UpdateVirtualMachine 

func UpdateVirtualMachine(ctx context.Context, client azclient.VirtualMachinesAPI, rg, vmName string, updates armcompute.VirtualMachineUpdate) error

func ZeroAKSMachineTimestamp added in v1.7.0 

func ZeroAKSMachineTimestamp() time.Time

Types

type AKSMachinePromise added in v1.7.0 

type AKSMachinePromise struct {
	AKSMachineTemplate *armcontainerservice.Machine
	AKSMachineName     string
	InstanceType       *corecloudprovider.InstanceType // Despite the reference nature, this is guaranteed to exist
	CapacityType       string
	Zone               string

	AKSMachineID               string
	AKSMachineNodeImageVersion string
	VMResourceID               string
	// contains filtered or unexported fields
}

Notes on terminology: An "instance" is a remote object, created by the API based on the template. A "template" is a local struct, populated from Karpenter-provided parameters with the logic further below. A "template" shares the struct with an "instance" representation. But read-only fields may not be populated. Ideally, the types should have been separated to avoid making cross-module assumption of the existence of certain fields.

TODO: Consider extracting the template-related fields (AKSMachineTemplate, AKSMachineName, InstanceType, CapacityType, Zone, AKSMachineID, AKSMachineNodeImageVersion, VMResourceID) into a dedicated struct (e.g., AKSMachineDetails or AKSMachineTemplateInfo). This would clarify the relationship between the promise fields and functions like BuildNodeClaimFromAKSMachineTemplate, as well as reduce the number of loose arguments passed around. More discussion: https://github.com/Azure/karpenter-provider-azure/pull/1197#discussion_r2482957255

func NewAKSMachinePromise added in v1.7.0 

func NewAKSMachinePromise(
	providerRef AKSMachineProvider,
	aksMachineTemplate *armcontainerservice.Machine,
	waitFunc func() error,
	aksMachineName string,
	instanceType *corecloudprovider.InstanceType,
	capacityType string,
	zone string,
	aksMachineID string,
	aksMachineNodeImageVersion string,
	vmResourceID string,
) *AKSMachinePromise

func (*AKSMachinePromise) Cleanup added in v1.7.0 

func (p *AKSMachinePromise) Cleanup(ctx context.Context) error

func (*AKSMachinePromise) GetInstanceName added in v1.7.0 

func (p *AKSMachinePromise) GetInstanceName() string

func (*AKSMachinePromise) Wait added in v1.7.0 

func (p *AKSMachinePromise) Wait() error

type AKSMachineProvider added in v1.7.0 

type AKSMachineProvider interface {
	// BeginCreate starts the creation of an AKS machine instance.
	// Returns a promise that must be waited on to complete the creation.
	BeginCreate(ctx context.Context, nodeClass *v1beta1.AKSNodeClass, nodeClaim *karpv1.NodeClaim, instanceTypes []*corecloudprovider.InstanceType) (*AKSMachinePromise, error)
	// Update updates the AKS machine instance with the specified name. Uses ETag for optimistic concurrency control.
	// Return NodeClaimNotFoundError if not found.
	Update(ctx context.Context, aksMachineName string, aksMachine armcontainerservice.Machine, etag *string) error
	// Get retrieves the AKS machine instance with the specified AKS machine name. Return NodeClaimNotFoundError if not found.
	Get(ctx context.Context, aksMachineName string) (*armcontainerservice.Machine, error)
	// List lists all AKS machine instances in the cluster.
	List(ctx context.Context) ([]*armcontainerservice.Machine, error)
	// Delete deletes the AKS machine instance with the specified name. Return NodeClaimNotFoundError if not found.
	Delete(ctx context.Context, aksMachineName string) error
	// GetMachinesPoolLocation returns the location of the AKS machines pool. The only reason this need to be exported is because armcontainerservice.Machine does not have the location field.
	GetMachinesPoolLocation() string
}

type DefaultAKSMachineProvider added in v1.7.0 

type DefaultAKSMachineProvider struct {
	// contains filtered or unexported fields
}

func NewAKSMachineProvider added in v1.7.0 

func NewAKSMachineProvider(
	ctx context.Context,
	azClient *azclient.AZClient,
	instanceTypeProvider instancetype.Provider,
	imageResolver imagefamily.Resolver,
	offeringsCache *cache.UnavailableOfferings,
	subscriptionID string,
	clusterResourceGroup string,
	clusterName string,
	aksMachinesPoolName string,
	aksMachinesPoolLocation string,
) *DefaultAKSMachineProvider

func (*DefaultAKSMachineProvider) BeginCreate added in v1.7.0 

func (p *DefaultAKSMachineProvider) BeginCreate(
	ctx context.Context,
	nodeClass *v1beta1.AKSNodeClass,
	nodeClaim *karpv1.NodeClaim,
	instanceTypes []*corecloudprovider.InstanceType,
) (*AKSMachinePromise, error)

BeginCreate creates an instance given the constraints. Note that the returned instance may not be finished provisioning yet. Errors that occur on the "sync side" of the VM create, such as BadRequest due to invalid user input, and similar, will have the error returned here. Errors that occur on the "async side" of the VM create (after the request is accepted) will be returned from AKSMachinePromise.Wait().

func (*DefaultAKSMachineProvider) Delete added in v1.7.0 

func (p *DefaultAKSMachineProvider) Delete(ctx context.Context, aksMachineName string) error

func (*DefaultAKSMachineProvider) Get added in v1.7.0 

func (p *DefaultAKSMachineProvider) Get(ctx context.Context, aksMachineName string) (*armcontainerservice.Machine, error)

ASSUMPTION: the AKS machine will be in the current p.aksMachinesPoolName. Otherwise need rework to pass the pool name in.

func (*DefaultAKSMachineProvider) GetMachinesPoolLocation added in v1.7.0 

func (p *DefaultAKSMachineProvider) GetMachinesPoolLocation() string

func (*DefaultAKSMachineProvider) List added in v1.7.0 

func (p *DefaultAKSMachineProvider) List(ctx context.Context) ([]*armcontainerservice.Machine, error)

func (*DefaultAKSMachineProvider) SetFallbackAKSMachinePollerOptions 

func (p *DefaultAKSMachineProvider) SetFallbackAKSMachinePollerOptions(opts aksmachinepoller.Options)

SetFallbackAKSMachinePollerOptions overrides the default GET-based poller options (used in tests to avoid real polling delays). "Fallback" because the GET poller is used when the SDK poller is nil (batch code path).

func (*DefaultAKSMachineProvider) Update added in v1.7.0 

func (p *DefaultAKSMachineProvider) Update(ctx context.Context, aksMachineName string, aksMachine armcontainerservice.Machine, etag *string) error

type DefaultVMProvider added in v1.6.6 

type DefaultVMProvider struct {
	// contains filtered or unexported fields
}

func NewDefaultVMProvider added in v1.6.6 

func NewDefaultVMProvider(
	azClient *azclient.AZClient,
	instanceTypeProvider instancetype.Provider,
	launchTemplateProvider *launchtemplate.Provider,
	loadBalancerProvider *loadbalancer.Provider,
	networkSecurityGroupProvider *networksecuritygroup.Provider,
	offeringsCache *cache.UnavailableOfferings,
	location string,
	resourceGroup string,
	subscriptionID string,
	provisionMode string,
	diskEncryptionSetID string,
	env *auth.Environment,
) *DefaultVMProvider

func (*DefaultVMProvider) BeginCreate added in v1.6.6 

func (p *DefaultVMProvider) BeginCreate(
	ctx context.Context,
	nodeClass *v1beta1.AKSNodeClass,
	nodeClaim *karpv1.NodeClaim,
	instanceTypes []*corecloudprovider.InstanceType,
) (*VirtualMachinePromise, error)

BeginCreate creates an instance given the constraints. instanceTypes should be sorted by priority for spot capacity type. Note that the returned instance may not be finished provisioning yet. Errors that occur on the "sync side" of the VM create, such as quota/capacity, BadRequest due to invalid user input, and similar, will have the error returned here. Errors that occur on the "async side" of the VM create (after the request is accepted) will be returned from VirtualMachinePromise.Wait().

func (*DefaultVMProvider) Delete added in v1.6.6 

func (p *DefaultVMProvider) Delete(ctx context.Context, resourceName string) error

func (*DefaultVMProvider) DeleteNic added in v1.6.6 

func (p *DefaultVMProvider) DeleteNic(ctx context.Context, nicName string) error

func (*DefaultVMProvider) Get added in v1.6.6 

func (p *DefaultVMProvider) Get(ctx context.Context, vmName string) (*armcompute.VirtualMachine, error)

func (*DefaultVMProvider) GetNic added in v1.6.6 

func (p *DefaultVMProvider) GetNic(ctx context.Context, rg, nicName string) (*armnetwork.Interface, error)

func (*DefaultVMProvider) List added in v1.6.6 

func (p *DefaultVMProvider) List(ctx context.Context) ([]*armcompute.VirtualMachine, error)

func (*DefaultVMProvider) ListNics added in v1.6.6 

func (p *DefaultVMProvider) ListNics(ctx context.Context) ([]*armnetwork.Interface, error)

ListNics returns all network interfaces in the resource group that have the nodepool tag

func (*DefaultVMProvider) Update added in v1.6.6 

func (p *DefaultVMProvider) Update(ctx context.Context, vmName string, update armcompute.VirtualMachineUpdate) error

Update updates the VM with the given updates. If Tags are specified, the tags are also updated on the associated network interface and VM extensions. Note that this means that this method can fail if the extensions have not been created yet. It is expected that the caller handles this and retries the update to propagate the tags to the extensions once they're created.

type Promise added in v1.6.6 

type Promise interface {
	// Cleanup removes the instance from the cloud provider.
	Cleanup(ctx context.Context) error
	// Wait blocks until the instance is ready.
	Wait() error
	// GetInstanceName returns the name of the instance. Recommended to be used for logging only due to generic nature.
	GetInstanceName() string
}

Intended for lifecycle handling on the higher abstractions.

type Resource 

type Resource = map[string]interface{}

func GetResourceData 

func GetResourceData(ctx context.Context, client azclient.AzureResourceGraphAPI, req arg.QueryRequest) ([]Resource, error)

Queries Azure Resource Graph using Resources() and returns a list of all pages of data.

type VMProvider added in v1.6.6 

type VMProvider interface {
	BeginCreate(context.Context, *v1beta1.AKSNodeClass, *karpv1.NodeClaim, []*corecloudprovider.InstanceType) (*VirtualMachinePromise, error)
	Get(context.Context, string) (*armcompute.VirtualMachine, error)
	List(context.Context) ([]*armcompute.VirtualMachine, error)
	Delete(context.Context, string) error
	Update(context.Context, string, armcompute.VirtualMachineUpdate) error
	GetNic(context.Context, string, string) (*armnetwork.Interface, error)
	DeleteNic(context.Context, string) error
	ListNics(context.Context) ([]*armnetwork.Interface, error)
}

type VirtualMachinePromise added in v1.4.0 

type VirtualMachinePromise struct {
	VM       *armcompute.VirtualMachine
	WaitFunc func() error
	// contains filtered or unexported fields
}

func (*VirtualMachinePromise) Cleanup added in v1.6.6 

func (p *VirtualMachinePromise) Cleanup(ctx context.Context) error

func (*VirtualMachinePromise) GetInstanceName added in v1.6.6 

func (p *VirtualMachinePromise) GetInstanceName() string

func (*VirtualMachinePromise) Wait added in v1.4.0 

func (p *VirtualMachinePromise) Wait() error

Source Files

View all Source files

Directories

Path Synopsis
Package aksmachinepoller provides a GET-based poller for tracking individual AKS machine provisioning status by polling GET machine until terminal state.
 Package aksmachinepoller provides a GET-based poller for tracking individual AKS machine provisioning status by polling GET machine until terminal state.
Context utilities for passing batch metadata through the call stack.
 Context utilities for passing batch metadata through the call stack.

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.