Documentation
¶
Index ¶
- type Decider
- type DeciderSpec
- type DeciderStatus
- type MultiScaler
- func (m *MultiScaler) Create(_ context.Context, decider *Decider) (*Decider, error)
- func (m *MultiScaler) Delete(_ context.Context, namespace, name string)
- func (m *MultiScaler) Get(_ context.Context, namespace, name string) (*Decider, error)
- func (m *MultiScaler) Inform(event types.NamespacedName) bool
- func (m *MultiScaler) Poke(key types.NamespacedName, stat metrics.Stat)
- func (m *MultiScaler) Update(_ context.Context, decider *Decider) (*Decider, error)
- func (m *MultiScaler) Watch(fn func(types.NamespacedName))
- type ScaleResult
- type UniScaler
- type UniScalerFactory
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type Decider ¶
type Decider struct {
metav1.ObjectMeta
Spec DeciderSpec
Status DeciderStatus
}
Decider is a resource which observes the request load of a Revision and recommends a number of replicas to run. +k8s:deepcopy-gen=true
func (*Decider) DeepCopy ¶
DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new Decider.
func (*Decider) DeepCopyInto ¶
DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
type DeciderSpec ¶
type DeciderSpec struct {
MaxScaleUpRate float64
MaxScaleDownRate float64
// The metric used for scaling, i.e. concurrency, rps.
ScalingMetric string
// The value of scaling metric per pod that we target to maintain.
// TargetValue <= TotalValue.
TargetValue float64
// The total value of scaling metric that a pod can maintain.
TotalValue float64
// The burst capacity that user wants to maintain without queuing at the POD level.
// Note, that queueing still might happen due to the non-ideal load balancing.
TargetBurstCapacity float64
// ActivatorCapacity is the single activator capacity, for subsetting.
ActivatorCapacity float64
// PanicThreshold is the threshold at which panic mode is entered. It represents
// a factor of the currently observed load over the panic window over the ready
// pods. I.e. if this is 2, panic mode will be entered if the observed metric
// is twice as high as the current population can handle.
PanicThreshold float64
// StableWindow is needed to determine when to exit panic mode.
StableWindow time.Duration
// ScaleDownDelay is the time that must pass at reduced concurrency before a
// scale-down decision is applied.
ScaleDownDelay time.Duration
// InitialScale is the calculated initial scale of the revision, taking both
// revision initial scale and cluster initial scale into account. Revision initial
// scale overrides cluster initial scale.
InitialScale int32
// Reachable describes whether the revision is referenced by any route.
Reachable bool
// ActivationScale is the minimum, non-zero value that a service should scale to.
// For example, if ActivationScale = 2, when a service scaled from zero it would
// scale up two replicas in this case. In essence, this allows one to set both a
// min-scale value while also preserving the ability to scale to zero.
// ActivationScale must be >= 2.
ActivationScale int32
}
DeciderSpec is the parameters by which the Revision should be scaled.
type DeciderStatus ¶
type DeciderStatus struct {
// DesiredScale is the target number of instances that autoscaler
// this revision needs.
DesiredScale int32
// ExcessBurstCapacity is the difference between spare capacity
// (how much more load the pods in the revision deployment can take before being
// overloaded) and the configured target burst capacity.
// If this number is negative: Activator will be threaded in
// the request path by the PodAutoscaler controller.
ExcessBurstCapacity int32
}
DeciderStatus is the current scale recommendation.
type MultiScaler ¶
type MultiScaler struct {
// contains filtered or unexported fields
}
MultiScaler maintains a collection of UniScalers.
func NewMultiScaler ¶
func NewMultiScaler(
stopCh <-chan struct{},
uniScalerFactory UniScalerFactory,
logger *zap.SugaredLogger) *MultiScaler
NewMultiScaler constructs a MultiScaler.
func (*MultiScaler) Delete ¶
func (m *MultiScaler) Delete(_ context.Context, namespace, name string)
Delete stops and removes a Decider.
func (*MultiScaler) Inform ¶
func (m *MultiScaler) Inform(event types.NamespacedName) bool
Inform sends an update to the registered watcher function, if it is set.
func (*MultiScaler) Poke ¶
func (m *MultiScaler) Poke(key types.NamespacedName, stat metrics.Stat)
Poke checks if the autoscaler needs to be run immediately.
func (*MultiScaler) Watch ¶
func (m *MultiScaler) Watch(fn func(types.NamespacedName))
Watch registers a singleton function to call when DeciderStatus is updated.
type ScaleResult ¶ added in v0.15.0
type ScaleResult struct {
// DesiredPodCount is the number of pods Autoscaler suggests for the revision.
DesiredPodCount int32
// ExcessBurstCapacity is computed headroom of the revision taking into
// the account target burst capacity.
ExcessBurstCapacity int32
// ScaleValid specifies whether this scale result is valid, i.e. whether
// Autoscaler had all the necessary information to compute a suggestion.
ScaleValid bool
}
ScaleResult holds the scale result of the UniScaler evaluation cycle.
type UniScaler ¶
type UniScaler interface {
// Scale computes a scaling suggestion for a revision.
Scale(*zap.SugaredLogger, time.Time) ScaleResult
// Update reconfigures the UniScaler according to the DeciderSpec.
Update(*DeciderSpec)
}
UniScaler records statistics for a particular Decider and proposes the scale for the Decider's target based on those statistics.
func New ¶
func New( reporterCtx context.Context, namespace, revision string, metricClient metrics.MetricClient, podCounter resources.EndpointsCounter, deciderSpec *DeciderSpec) UniScaler
New creates a new instance of default autoscaler implementation.
type UniScalerFactory ¶
UniScalerFactory creates a UniScaler for a given PA using the given dynamic configuration.
Source Files